Posts from the Compucell3D AllAnswered Forum.

Created: April 2019 jps

Q1: Which operating system CompuCell3D users use?

content:

We would like to know about operating system CompuCell3D users use. This will allow us to support and development CompuCell3D more efficiently.

Answer:

content:
Microsoft Windows

(Includes Windows 10, Windows 7 all Microsoft platforms)

Answer:

content:
Apple MacOS

( All Apple products)

Answer:

content:

Linux

( All Linux platforms and flavors e.g. Ubuntu, RedHat, Mint)

Q2: What is CompuCell3D?

content:
CompuCell3D is an open-source simulation environment for multi-cell, single-cell-based modeling of tissues, organs and organisms.

It uses Cellular Potts Model to model cell behavior. For technical details about modeling please refer Multi-Scale Modeling of Tissues Using CompuCell3D paper.

For more information and tutorials please visit CompuCell3D website.

This project is funded by the NIH and EPA.

Q3: How to use this forum?

content:
This CompuCell3D forum should only be used for following purposes-

1. Asking solution for an issue 
2. Suggesting or discussing a new feature 

Q4: How can I know the total count of cells for specific cell type?

content:
system: Windows 7
CC3D Version: 3.7.5

Answer:

content:

Assuming you have two cell types - Condensing and NonCondensing you could do the following:

    def step(self, mcs):

        print 'num condensing=', len(self.cellListByType(self.CONDENSING))
        print 'num noncondensing=', len(self.cellListByType(self.NONCONDENSING))

Alternatively, instead of using self.CONDENSING and self.NONCONDENSING you may use cell type ids directly (I assume Condensing has cell type id =1 and NonCondensing has cell type id =2):

    def step(self, mcs):

        print 'num condensing=', len(self.cellListByType(1))
        print 'num noncondensing=', len(self.cellListByType(2))

Q5: How can set a layer of cells that get together like a thin film?

content:
I want to use some cells set a thin film but when model starts running those cells won't stay together and disperse. I know "Cell Type" plugin that has a function named  "freeze". But the "Freeze" function will not let cells to be altered by pixel copies. I wonder if there are some other methods which can keep cells stay together like a film during simulation.

Answer:

content:
"Playing" with the contact energy may help. If the contact energy between cells of the same type is decreased, affinity is increased, so dispersion is less likely, though they may clump together instead. For a layer of cells (L) sitting on top of another group of cells (S), you want the contact energies to be S-S < S-L < L-L < L-medium I believe.

Answer:

content:
There are a couple ways to make cells form a layer.
1. You can have the cells stick to the wall of the bounding box (but then you can only access one face of the cell layer).
2. You could make a curved cell layer ala the cell sorting demos where your cell layer is actually spherical (the "noncondensing" cells in the cell sort demos).
3. You can use compartmentalized cells that have basal, apical and lateral faces defined.
4. You can add distance constraints between cells using "Elasticity Plugin" or "FocalPointPlasticity Plugin". Constraints on contact neighbors keep the sheet intact. If the distance constraints go beyond first neighbors then the constraint will tend to keep the sheet flat.

You might take a look at Biofilms in 2D https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2547990/  SIMULATION OF SINGLE-SPECIES BACTERIAL-BIOFILM GROWTH USING THE GLAZIER-GRANER-HOGEWEG MODEL AND THE COMPUCELL3D MODELING ENVIRONMENT

Are you working in 2D or 3D?

Comment:

content:
I agree with Jim's 4. suggestion. This would be the most robust way to have cells stay together. Relying purely on contact energy may not be sufficient. It would help us if you could post the geometry of the tissue you are trying to model. 

Comment:

content:
i just want to make cells form a layer to simulate a rectangular hydrogel film，so that i can simulate the mechanical properties of hydrogel in simulation。

Answer:

content:
Attached is a zip file with a CC3D model of a layer of cells in 2D (basically a line of cells). The line is maintained using focal point plasticity links. The sim first creates the line of cells, it then displaces the central cell upwards and the end cells downward showing the sheet stays intact (though holes open transiently). It then pulls the ends together into a circle.

It should be possible to convert into a 3D model.

CC3D files: File attached: CellsInLayer_FPP.zip (1.05 MB)

An animation of the output: File attached: output_ObCKSA.mp4 (141.98 KB)

attachments(redirected):
.\CC3D_allanswered_linked_files\CellsInLayer_FPP.zip
.\CC3D_allanswered_linked_files\output_ObCKSA.mp4

Comment:

content:
If the hydrogel is coating some object (e.g. tissue) then all you need to do is to create a blob of tissue, cover it with gel and choose appropriate adhesion (ContactPlugin) coefficients

Hydrogel.xml

<CompuCell3D Revision="20170429" Version="3.7.6">
   
   <Potts>
      <Dimensions x="100" y="100" z="1"/>
      <Steps>10000</Steps>
      <Temperature>20.0</Temperature>
      <NeighborOrder>1</NeighborOrder>
   </Potts>
   
   <Plugin Name="CellType">
      <CellType TypeId="0" TypeName="Medium"/>
      <CellType TypeId="1" TypeName="Gel"/>
      <CellType TypeId="2" TypeName="Tissue"/>
   </Plugin>
   
   <Plugin Name="Volume"/>
   
   <Plugin Name="CenterOfMass"/>
   
   <Plugin Name="Contact">
      <Energy Type1="Medium" Type2="Medium">10.0</Energy>
      <Energy Type1="Medium" Type2="Gel">25.0</Energy>
      <Energy Type1="Medium" Type2="Tissue">25.0</Energy>
      <Energy Type1="Gel" Type2="Gel">12.0</Energy>
      <Energy Type1="Tissue" Type2="Tissue">5.0</Energy>
      <Energy Type1="Gel" Type2="Tissue">10.0</Energy>
      <NeighborOrder>3</NeighborOrder>
   </Plugin>
   
</CompuCell3D>​

import sys
from os import environ
from os import getcwd
import string

sys.path.append(environ["PYTHON_MODULE_PATH"])


import CompuCellSetup


sim,simthread = CompuCellSetup.getCoreSimulationObjects()
            
# add extra attributes here
            
CompuCellSetup.initializeSimulationObjects(sim,simthread)
# Definitions of additional Python-managed fields go here
        
#Add Python steppables here
steppableRegistry=CompuCellSetup.getSteppableRegistry()
        
from HydrogelSteppables import HydrogelSteppable
steppableInstance=HydrogelSteppable(sim,_frequency=1)
steppableRegistry.registerSteppable(steppableInstance)
        
CompuCellSetup.mainLoop(sim,simthread,steppableRegistry)
        
        ​

from PySteppables import *
import CompuCell
import sys

class HydrogelSteppable(SteppableBasePy):    

    def __init__(self,_simulator,_frequency=1):
        SteppableBasePy.__init__(self, _simulator, _frequency)
    def start(self):
        
        cell_size = 5
        rows = 10
        cols = 10
        
        blob_pos_vec = [20, 20, 0]
        
        for r in range(rows):
            for c in range(cols):            
                new_cell = self.newCell(self.TISSUE) 
                x_pos = r*cell_size + blob_pos_vec[0]
                y_pos = c*cell_size + blob_pos_vec[1]
                self.cellField[x_pos:x_pos+cell_size, y_pos:y_pos+cell_size, 0] = new_cell
                if r==0 or c==0 or r==rows-1 or c==cols-1:
                    new_cell.type = self.GEL    

        #setting up volume constraint
        for cell in self. cellList:
            cell.targetVolume = 25
            cell.lambdaVolume = 2.0
            if cell.type == self.GEL:
                cell.targetVolume = 30
                cell.lambdaVolume = 2.0
                
        
    def step(self,mcs):        
        pass

​

Comment:

content:
Deeply thanks for your help，and sorry for my unclear expression. The tissue is not coated by hydrogel but cultured on hydrogel. Assuming a simulation has 2 units of z axis, one unit is a layer of hydrogel and another unit is a 2d tissue cultured on that hydrogel. Can this be achieved in cc3d？  

Comment:

content:
thanks for your help

Comment:

content:
Thanks for your help，my previous expression is not clear. The supplementary explanation is under Maciek Swat’s comment. 

Q6: In a 3D model - How can I let cells grow along a wall?

content:
In my model, I set a wall and let cells grow on it. If I don't set specific condition, cell will growing to a lump. I tried to use plugin 'Focal Point Plasticity' to restrain cell growth along the wall but it's not effective enough. I wonder whether I should use this plugin according with in vitro anchorage-dependent cell growth status？

Answer:

content:
FPP will do it but you might first try:

1. Have the cells sticker to the wall than to themselves. Use low stickyness to medium, with medium-wall being the least sticky of all. The magnitudes will need to be adjusted differently in 2D vs. 3D.
(assuming wall is fixed so wall-wall adhesion doesn't matter)
-3 wall-cell
-2 cell-cell
0 cell-medium
2 wall-medium
0 medium-medium
Note that the adhesion energy will tend to flatten the cells out. We can take advantage of that in the next suggestion.

2. When dividing cells divide along the major axis. This will usually keep both the parent and daughter cells in contact with the surface. Search for "major axis" in the documentation, it is one of the options for the mitosis plugin and steppable.

Answer:

content:
If you are still having problems we should be able to come up with a demo of using FPP to help keep cells sticking to a wall or surface.

Comment:

content:
I used FPP and the method above together to help cells stick to wall. It seems to have worked, but I found that cell is fragmented after a few steps even if I used Connectivity Plugins.

Answer:

content:
I am not sure if this feature already exist in CC3D but ideally what you should be able to do is to use elongation plugin - it elongates cells AND specify the direction of elongation -  i.e. a vector of a major axis of the cell. There is a plugin called "OrientedGrowth" which probably does what you want. We do not have examples of how to use this plugin in our demo suite but I can reach out to a person who contributed this plugin and find out. 

Another option to explore is to use compartmentalized cells (i.e. your biological cell is a collection of CompuCell3D cells - compartments) and then you could use Focal Point Plasticity to connect the compartments together. Of course the compartments would still remain "blobs" but you could get elongated cells this way.

Comment:

content:
Could you post your simulation showing how connectivity plugin leads to fragmentation? This should not happen so we would like to investigate the cause

Answer:

content:
This demo was contributed by Jeremy Fisher (https://cc3dadvancedtuts.wordpress.com/) who also coded up the OrientedGrowth plugin. Take a looks and see if this looks closer to what you are looking for

File attached: OrientedGrowth.zip (2 KB)

attachments(redirected):
.\CC3D_allanswered_linked_files\OrientedGrowth.zip

Answer:

content:
Oriented growth should impose an arbitrarily large energy penalty on growth perpendicular to the wall, which I believe this would address the issue. However, it only supports 2D simulations as of the moment.

It wouldn't be too difficult to add support for 3D simulations. Is this an approach you would be interested in?

Comment:

content:
the problem of fragmentation is fixed, Connectivity Plugin's invalid maybe because of I set the value of Penalty is out of threshold. Now the model is basically achieved expectations. But when cells grow to relatively high density, it's hard to avoid cell overlap and some cells will tend to growing away from wall. So I wonder how can i know the cell neighbor's cell type that can help me to eliminate cells which do not contact the wall. The CompuCell3D 3.7.4 Reference Manual only tell us how to get cell neighbor' id but not their type.

Comment:

content:
You can visit all cells of a particular type and then check the contact area for that cells neighbors of particular types with something like;

for cell in self.cellListByType(self.CULTUREDCELLS): ### list of  cell types (capitalized)
    cell_neighbors = self.getCellNeighborDataList(cell)  ### list of this cell’s neighbors
    neighbor_count_by_type = cell_neighbors.neighborCountByType()
        ### neighbor_count_by_type is a hash indexed by cell type IDs,
        ### so neighbor_count_by_type{1} gives the number of neighbors of cell type 1
        ### if no neighbors of a particular type you could do something here

    if  (cell_neighbors.commonSurfaceAreaWithCellTypes([self.HYDROGEL]) < (cellWidth**2/4.)):
        # this cell has very little contact with hydrogel so kill it
        cell.targetVolume  = 0.
        cell.targetSurface = 0.
        cell.lambdaVolume  = 0.01  # small lambda's so the cell shrinks slowly
        cell.lambdaSurface = 0.01
        print("KILL {}: {}, tv={} ts={} v={} s={}".format(cell.id,cell.type,cell.targetVolume,cell.targetSurface,cell.volume,cell.surface))

In the main xml you need;
<Plugin Name="NeighborTracker">
   <!-- Module tracking neighboring cells of each cell -->
</Plugin>

Comment:

content:
It works, truely thanks for your help.

Q7: Can there be two different cell types simultaneously residing in the same pixel?

content:
Hi,

So what I observed from the movement of CC3D generalized cells is that they all move around each other, and not over each other. Is is possible to implement, say the phenomena where tumor cell moves along collagen fibers in the extracellular matrix? From my current knowledge of CompuCell3D, in order for this phenomenon to be possible, two different cell types must be on the same pixel at the same simulation step. Is this possible in CompuCell3D simulation? Also, in my simulation, I treat fibers as discrete generalized cells (I also freeze them) instead of chemical field. 

Sincerely, 

Answer:

content:
CC3D can't have the same pixel as part of two cells, indeed biology and physics don't allow that either. The only time you would need that is if you model a 3D system (e.g., real life) in 2D (not real life). So, if you do your model in 3D there isn't any problem of overlapping cells  and one cell can crawl along another cell (or pseudo cell).

If you really want 2D (perhaps because of the much faster calculation speed) then you can;
1. Have a cell crawl along one side of the fiber (easy), cells can't cross the fiber and you'll be limited to a simple fiber topology (what does the cell do when it reaches a point where two fibers cross?).
2. With a fair amount of fiddling get it to cross over the fiber by basically taking control of the identity of individual pixels (you can arbitrarily assign any pixel to any cell in Python).
3. You could treat the crawling cell as two separate cells, one on each side of the fiber, then link the two across the fiber with a focal point plasticity (or a spring) link.
4. If you really want 2D and a complex fiber topology with things like fibers crossing, then treating the fiber as a field is probably the way to go. There are two ways to do this;
4a. As a CC3D field, you can define a field that doesn't diffuse (set the diffusion constant to zero) and you can directly poke values into a field from python so you can generate an arbitrarily complex layout of fibers. This will work but even with the diffusion constant set to zero I believe the diffusion calculations will still be done and that might slow things down somewhat.
4b. Create a python (or numpy) 2D array with the same dimensions as the CC3D model and keep the fiber network in that array. No diffusion calculations are done. You can then use python to have that array interact with your cells by having the cell movement being a function of the interaction between the CC3D cell pixels and the separate matrix mapping the fibers.

In 3D things would be more straight forward, but also much slower.

Think about how you want to do things and let us know if you need help.

Answer:

content:
In principle we could implement this feature but it would take a bit of code refactoring. Modeling ECM is somewhat challenging because the granularity of fibers is much finer than the way we represent cells on a fairly coarse lattice. I think that modelling ECM using external module where we model fibers as a e.g. vector field could be a better approach. You could "translate" deformation in the cell field into forces acting on fibers and vice-versa, translate stresses in the ECM fields into forces acting on  cells ( in the the CC3D cell field). The "integration" of the ECM model with the cell-based model could fairly easily be done in Python . The advantage of such approach is that you would not have to modify existing CCC3D code but rather add external ECM modeling module that could be more sophisticated than anything you could currently mimic in CC3D

Q8: Page 11, Section 3.2 of Compucell 3D Manual and Tutorial, Version 3.6.0

content:
On page 11, Section 3.2 of Compucell 3D Manual and Tutorial, Version 3.6.0, in the first paragraph, it is written:

"During each index-copy attempt, we select a target pixel, i, randomly from the cell lattice, and then randomly select one of its neighboring pixels, i, as a source pixel."

Should the second "i" be a j?

-Scott

Answer:

content:
Yes :)
I wonder how many version of the manual this error traces back in?

Answer:

content:

Actually, the mentioned neighboring pixel is not $i$i  it's  $i'$i' ( i-prime or I-dash). Snippet from the manual-

Unfortunately, it's little confusing given vector bar above . Let's see if we can use  $j$j instead  $i'$i' in next version of manual. Thank you Scott for pointing it out. 

attachments(redirected):
.\CC3D_allanswered_linked_files\Screen Shot 2017-10-24 at 2.04.12 PM.png

Answer:

content:
Thank you, Anwar, that makes more sense.

I see the latest version is 3.7.4 on the CC3D website. 

Q9: Open VTK files for visualization in other software

content:
Dear CC3D users,

Are there any other visualization software that can be used to visualize the trajectory output from CC3D and make fancy movies? I've tried VMD to open the .vtk files and it seems to be work but probably some tweaks need to be done to get the correct visuals. Anyone has experience with that?

Answer:

content:
VTK is kind of a difficult format. Besides VMD there are also (at least);

• Paraview http://www.paraview.org
• VisIt https://wci.llnl.gov/simulation/computer-codes/visit/
• VTKDesigner https://sourceforge.net/projects/vtkdesigner/

Comment:

content:
Thank you very much for the advice! I'm testing out Paraview together with output from the cellsort_3D demo. I immediately ran into a problem when I tried to visualize by celltype, I get the error of

ERROR: In /Users/kitware/dashboards/buildbot-slave/8275bd07/build/superbuild/paraview/src/VTK/Rendering/Volume/vtkGPUVolumeRayCastMapper.cxx, line 353
vtkOpenGLGPUVolumeRayCastMapper (0x7ff64c9b5480): scalar of type VTK_CHAR is not supported because this type is platform dependent. Use VTK_SIGNED_CHAR or VTK_UNSIGNED_CHAR instead.​

find . -type f -name 'cellsort_3D_cc3d_*' -exec sed -i 's/char/unsigned_char/g' {} \;​

Answer:

content:
Yes, manually changing the celltype data type in the VTK file(s) appears to work, for example;

Comment:

content:
I'm using the latest version of Paraview 5.4.1 on MacOSX Sierra.

Q10: Import CompuCell3D 3.3.1 into CompuCell3D 3.7.6

content:
Hi, I'm a student and I want to import the code of the aticle: http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0007190 into new CompuCell3D, but in this code base there is not a .cc3d file. How can I do?

Answer:

content:
Basically .cc3d is a XML file enlisting all related files and resources for the simulation. I believe you can generate a .cc3d file manually referring to demos in CompuCell3D 3.7.6.

Please let us know if you need further help with that.

Answer:

content:
Typically this is what a .cc3d file looks like

<Simulation version="version number">
<XMLScript Type="XMLScript">XML_File_Name</XMLScript>
<PythonScript Type="PythonScript">Python_File_Name</PythonScript>
<Resource Type="Python">Python_Steppable_File_Name</Resource>
</Simulation>

You might need to add additional location information in the file names if they are in different directories.

Answer:

content:
The easiest thing to do is to just create a new model in Twedit++ (basically just pick a few items at random), then cut and paste the XML and python from the published model into the appropriate files in your newly created model.

Comment:

content:
Hi, I try but do not work, the error in console is:

FILE NAME= Traceback (most recent call last):
  File "/usr/lib/compucell3d/Twedit++/Plugins/PluginCCDProject.py", line 2443, in __openCC3DProject
    self.showOpenProjectDialogAndLoad(currentFilePath)
  File "/usr/lib/compucell3d/Twedit++/Plugins/PluginCCDProject.py", line 2430, in showOpenProjectDialogAndLoad
    print "FILE NAME=",fileName
UnicodeEncodeError: 'ascii' codec can't encode character u'\xe0' in position 34: ordinal not in range(128)​

<Simulation version="3.6.2">
   <XMLScript Type="XMLScript">Simulation/XML_Template_TumorVasc3D_Movie.xml</XMLScript>
   <PythonScript Type="PythonScript">Simulation/angio_growth.py</PythonScript>
   <Resource Type="Python">Simulation/angio_growth_steppables.py</Resource>
   <Resource Type="Python">Simulation/angio_growth_plugins.py</Resource>
</Simulation>
​

Comment:

content:
I try with this method and I have this error in console:

projectFullPath= /home/nic/CompucellAutomaticProject/CompucellAutomaticProject.cc3dNEW SIMULATION reply=NEWSIMULATION/home/nic/CompucellAutomaticProject/CompucellAutomaticProject.cc3d starting CC3D READING RESPONSE CC3D SENDER self.socket.bytesAvailable()= 168 self.nextBlockSize= 166 self.socket.bytesAvailable()= 166  messageType= NEWSIMULATION self.errorConsole.playerMainWidget= <Plugins.ViewManagerPlugins.SimpleTabView.SimpleTabView object at 0x7f6c1b500770>

Comment:

content:
I believe this error is due to encoding mismatch with value in string 'fileName'. I suspect this occurred due to the .cc3d file is not saved with UTF-8 encoding.

Can you try after saving it with UTF-8 encoding? Google how to save it in UTF-8 encoding with whichever text editor you are using. Instructions for notepad++ are here.

Comment:

content:
Hi, I have solved some prolems, but now I have this:

messageType= NEWSIMULATION
self.errorConsole.playerMainWidget= <Plugins.ViewManagerPlugins.SimpleTabView.SimpleTabView object at 0x7f62a0e80a68>
processIncommingSimulation =  /home/mint/compuCellWorkspace/NewSimulation/NewSimulation.cc3d  _stopCurrentSim= True
GOT NEW SIMULATION  /home/mint/compuCellWorkspace/NewSimulation/NewSimulation.cc3d
NEWSIMULATIONRECEIVED SIMULATION NAME SENT SUCCESFULLY​

Q11: what is the meaning of contact energy's Positive value and Negative value

content:
Assume there are two cells are contacted，and set contact energy J to a negative value. Does it means those two cells will definitely stay together ， if set J to a positive value does it means those two cells possible to separate and its probability is decided by the value of J.

Answer:

content:
Usually, besides the j value between a pair of cell types you also need to take into account the j values for the two cell types with medium.

Besides that, you always need to take into account all possible interactions. So even if a particular pair of cells has a j value of -10 and a large contact area (contact energy is J*contact_area) that contact could still be displaced if either of the two cells has a higher affinity (more negative J) for some other cell type and/or the other cell type is more flexible and can acquire a larger contact surface area.

So, in general, you need to take into account all possible pairs of interactions; for N cell types (and assuming Medium is one of those N types) there are N^2/2 + N/2 = (N/2)(N+1) possible interactions. For 3 cell types there are 6 interaction, for 4 cell types there are 10 interactions, ...

Beyond the J values the actual interaction energy will also be dependent on how stiff the surface of the cells is. With a very low LambdaSurface and/or large target surface area, cells can create extra surface in order to reap the reward of a large contact surface if that lowers contact energy.

Finally, you need to remember that movement of a CC3D configuration to a lower energy configuration can only occur if a suitable pathway exists. In other words, it is possible to trap a CC3D system in an energetically unfavorable state if there is no reasonably low energy pathway to a lower energy state. (That is also true of biology, biological systems are trapped in high energy states (life) since a low energy pathway to lower energy states (death) isn't always readily available.)

Comment:

content:
Regardless of other conditions，does it means the relationship of contact strength is J=-10>J=-5>J=0>J=5>J=10？
And the meaning of J=0 is confused for me.

Comment:

content:
Yes, that is correct the j=-10 interaction is stronger than the j=10 one. A j of zero just means there is no contribution to the systems total energy due to that particular interaction. If all your j's are either (for example) +10 or zero then the system will try to get all the contacts to be of the zero type. If all your j's are -10 and zero then the system will try to get all the contacts to be of the -10 type.

Comment:

content:
Hi, I am new in CC3D, can we have a communication in Chinese? My QQ is 1775755174

Q12: Drawing 3D structures in CellDraw

content:
Dear Admins,

I took a look at the CellDraw manual and pottered around the buttons in CellDraw itself, but I could not figure out how to draw 3D regions. Is this specified by the "cell size" column in the cell type table?

Also, I am trying to simulate simple cell movement on a solid substrate, is there an easy way to specify a smoothly increasing contact adhesion energy of the cells with the substrate, say increasing adhesion energy from left-most to right-most region of the substrate?

Answer:

content:
For your second question; one approach would be to subdivide the substrate into pseudo-cells and then assign individual adhesion energies to each substrate cell based on their position. The AdhesionFlex plugin allows you to assign individual adhesivities to particular cells. Check in your CC3D demos folder for
C:\CompuCell3D-64bit\Demos\PluginDemos\AdhesionFlexPython\
as an example.

Since your substrate pseudo-cellos probably should not move you can set all cells of that type to "frozen".

Answer:

content:
Is that the same thing as Piff Generator? I found a Piff drawing program for windows, but I can't recall the name. I'll check once I get access to my Windows laptop.

Comment:

content:
How do you install Celldraw? I have the CC3D_3.7.7 binary downloaded. I have the Celldraw 1.6 from GitHub. 

scottholmes src $ ./CellDraw_standalone.command
====> CellDraw working directory: /Users/scottholmes/CC3D_3.7.7/CellDraw/1.6.0/src
====> PYTHONLIB26: /System/Library/Frameworks/Python.framework/Versions/2.6
====> DYLD_LIBRARY_PATH: /Users/scottholmes/CC3D_3.7.7/CellDraw/1.6.0/src/../Frameworks
====> PYTHONPATH directory: /Users/scottholmes/CC3D_3.7.7/CellDraw/1.6.0/src/../Resources/site-packages:
====> PATH: /System/Library/Frameworks/Python.framework/Versions/2.6/bin:
====> PREFIX_CC3D: /Applications/CC3D_3.6.0_MacOSX106
./CellDraw_standalone.command: line 37: /usr/bin/pythonw2.6: No such file or directory
====> CellDraw 1.6.0 now starting from Python.
./CellDraw_standalone.command: line 41: /usr/bin/pythonw2.6: No such file or directory

Q13: Compilation instructions for CC3D on multi-core and GPU processors

content:
May I know whether there are specific instructions on how to compile parallel CC3D with nVidia GPUs on Ubuntu or any other Linux-based clusters?

Also, can the pre-compiled binaries for both Mac OSX and Ubuntu be run on multiple cores? I do not seem to be able to find an option to do that.

Answer:

content:

Answer:

content:
Can the pre-compiled binaries for both Mac OSX and Ubuntu be run on multiple cores? 
Actually CC3D needs to be instructed to use multiple cores it is quite simple all you need to do is to to include <Metadata> tag with specification of how many CPU's you want to use

 <CompuCell3D>
  <Metadata>
    <NumberOfProcessors>4</NumberOfProcessors>
    <DebugOutputFrequency>100</DebugOutputFrequency>
 </Metadata>

 <Potts>
   <Dimensions x="200" y="200" z="1"/>
   <Anneal>10</Anneal>
   <Steps>10000</Steps>
   <Temperature>10</Temperature>
   <Flip2DimRatio>1</Flip2DimRatio>
   <NeighborOrder>2</NeighborOrder>
 </Potts>
...
</CompuCell3D>

​

Are there are specific instructions to compile parallel CC3D with nVidia GPUs on Ubuntu or any other Linux-based clusters?

One thing to remember about GPU execution that it applies only to PDE solvers. The core Potts algorithm and all your Python Steppables will still run on CPU. So the bottom line is that your entire simulation will be limited by the slowest component. Now, if you are running a lot of PDE solvers than using GPU's makes sense.

I have not tried compiling GPU modules on linux (did that for OSX and Windows only) but it really should be straightforward. All you need to do is to make sure that OpenCL is installed on your linux cluster and the CMake script will automatically detect the settings. If you look at the CMake configuration dialog below the two settings that decide if the CC3D will have GPU modules are "NO_OPENCL" option (make sure it is unchecked if you wan GPU modules) and the path to OPENCL library

Let me know if you need more help

attachments(redirected):
.\CC3D_allanswered_linked_files\cmake_cc3d.png

Answer:

content:
The compilation steps you should follow on Ubuntu are outlined on 
http://www.compucell3d.org/SrcBin/LinuxCompile
More compilation tutorial can be found on 
http://www.compucell3d.org/DeveloperZone

Answer:

content:
Thanks for the answers! I will try it out and let you guys know if I run into issues.

Comment:

content:
How many solvers do you mean by "a lot of PDE solvers"? I am running 5 PDE solvers - would it be better if I use GPUs?

Answer:

content:
you would be better off trying GPU solvers but for that you would need to install OpenCL drivers on your machine, provided your machine has a decent GPU . And to see nice speedups you need a decent GPU. Regular PDE solvers are actually quite fast. Another option to consider it to use external PDE-solver suite but this is more involved technically

Comment:

content:
Okay, thank you for your response! I'll try using GPUs.

Answer:

content:
I successfully built 3.7.4 on Red Hat Enterprise with OpenCL on a system with 4 Nvidia Tesla V100s.   I have confirmed that increasing the NumberOfProcessors  greater than 1 with OpenMP alone will yield blank results, but will work only by using OpenCL as well.  I ran benchmarks with 1, 8, 16, and 32 processors and DiffusionSolver_OpenCL, while faster than 1 processor alone by only 2.14 times, seems to be the bottleneck rather than the rest of the OpenMP processes, since 32 processors are only 4.4 times faster than 1 with the GPU.  So I am asking, given that I have a high end GPU, and CC3D knows nothing about my cell sizes relative to the block sizes it chooses, are there parameters I can change to make the diffusion solver work harder and faster on the GPU?  My test grid is 600x600x60, but will go larger in practice.

I also wonder if 3.7.9 has significant performance enhancements.  An important command parameter was removed from runScript since 3.7.4  that I need, and I run many simultaneous batch simulations headless on a large compute server with VTK/OSMesa doing the software rendering.

Richard,  USEPA

Q14: Segmentation fault

content:
File attached: NewSimulation.zip (814.89 KB)

In the project attached, if I open it in Twedit++ and click on CC3D Project>Open In Player, Compucell3D close itself and get a segmentation fault error.
(In the zip there is the complete error log in file "ERROE: SEGMENTATION FAULT.txt")
(I'm on linux.)

attachments(redirected):
.\CC3D_allanswered_linked_files\NewSimulation.zip

Comment:

content:
Yes. I am able to reproduce the same error on my machine with the given simulation. But not sure yet what exactly causing this error. I will try to find some time to debug this.

Answer:

content:
Are you by any chance using Linux installed under VirtualBox?

Answer:

content:
Ok so the problem with this simulation is mitosis plugin that has been deprecated for many years, as it had many drawbacks. These days the mitosis is handled more efficiently in the Python Steppable . Anyway, I am attaching two simulations  -  the original one that was posted above under NewSimulation.zip (VascularTumorOrig.zip) but without Mitosis and the simulation that follows new CompuCell3D API that implements model that is similar to the original one but not quite exactly (E599Lecture12AbbasPaperCancerSimulationforUpload.zip)

The original simulation witohut the mitosis plugin opens and you can run it but it will not produce anything interesting because cells will not divide. The updated simulation runs but we have not full tested if it yields the same results as the original simulation that was published. In any case the second simultion woudl be a great starting point for further explorations

Take a look at those simulations and see if indeed they open up and run on your machine

File attached: VascularTumorOrig.zip (815.27 KB)
File attached: E599Lecture12AbbasPaperCancerSimulationforUpload.zip (273.22 KB)

attachments(redirected):
.\CC3D_allanswered_linked_files\VascularTumorOrig.zip
.\CC3D_allanswered_linked_files\E599Lecture12AbbasPaperCancerSimulationforUpload.zip

Comment:

content:
It seems not work correctly, but start. Now I try to work about it, thanks!

Q15: Non-informative error when an SBML file doesn't exist

content:
If an SBML model file isn’t found the error message doesn’t say that, instead it says "NameError: global name 'os' is not defined".  (CC3D version 3.7.5)

Traceback (most recent call last):
  File "C:\CompuCell3D-64bit\pythonSetupScripts\CompuCellPythonSimulationNewPlayer.py", line 351, in <module>
    execfile(CompuCellSetup.simulationPaths.simulationPythonScriptName)
  File "C:\Users\James Sluka\Desktop\Small Code Projects\Multi_simple_liver\Liver_I_paper_model\cc3dCode\Simulation\PBPK
_SBML_MULTI.py", line 272, in <module>
    CompuCellSetup.mainLoop(sim,simthread,steppableRegistry)
  File "C:\CompuCell3D-64bit\pythonSetupScripts\CompuCellSetup.py", line 1761, in mainLoop
    return mainLoopNewPlayer(sim, simthread, steppableRegistry, _screenUpdateFrequency )
  File "C:\CompuCell3D-64bit\pythonSetupScripts\CompuCellSetup.py", line 1498, in mainLoopNewPlayer
    steppableRegistry.start()
  File "C:\CompuCell3D-64bit\pythonSetupScripts\PySteppables.py", line 1460, in start
    steppable.start()
  File "C:\Users\James Sluka\Desktop\Small Code Projects\Multi_simple_liver\Liver_I_paper_model\cc3dCode\Simulation\PBPK
_SBML_MULTISteppables.py", line 777, in start    self.addFreeFloatingSBML(_modelFile=modelFile,_modelName='PBPKWambaugh_cS',_stepSize=self.timeStepOfIntegration4,_in
itialConditions=initialConditions)
  File "C:\CompuCell3D-64bit\pythonSetupScripts\SBMLSolverHelper.py", line 184, in addFreeFloatingSBML
modelPathNormalized=os.path.abspath(os.path.join(self.simulator.getBasePath(),modelPathNormalized))
NameError: global name 'os' is not defined



Q16: How can I simulate the mitosis of a sphere cell?

content:
Hi, I want to simulate the mitosis process of a sphere cell on a substrate. How can I set the initial sphere shape with square or hexagonal lattice? Thank you~

Answer:

content:
To generate spherical cells, use the function self.cellField whose arguments are coordinates in the lattice space. If your radius is r, and center of the sphere at say (x0,y0,z0) the coordinates (x0-r,x0+r),(y0-r,y0+r) and (z0-r,z0+r) will span a bigger cube. Then the value of r can be used as a distance cutoff to determine which voxels will lie inside the boundary of the sphere. Example code:

#generate cell 
cella=self.newCell(some cell type);
#set up the center of sphere say (x0,y0,z0)
#set up the boundaries of the bigger cube
xi = x0 - r
xf = x0 + r
yi = y0-r
yf = y0+r
zi=z0-r
zf=z0+r
#loop over the points to determine boundaries of the circle
for xr in range(xi, xf):
  for yr in range(yi, yf):
   for zr in range(zi,zf):
    rd = sqrt((xr - x0) ** 2 + (yr - y0) ** 2+(zr-z0)**2)
     if (rd < r):
      self.cellField[xr, yr, zr] = cella

Comment:

content:
Thank you so much, I can set up the sphere shape in this way. But how to let the shape of the daughter cells keep sphere after the mitosis. I am going to simulate many times of mitosis.

Comment:

content:
Your daughter cells will tend to stay spherical if they have the proper target surface for their current and target volumes. So periodically change individual cell's target surface to;
      import math
      cell.targetSurface=((cell.targetVolume)**(1./3.))**2*4./3.*math.pi

The above is just a starting point for what you want to do. The surface of a "sphere" in a lattice representation is often a fair bit bigger than in a non-lattice situation. Often though if the target surface is too small that will tend to make the cell as spherical as possible. You then have to worry about the lambda for surface and volume and make sure that when you want to grow the cell's volume that lambdaVolume is big enough to overwhelm lambdaSurface amd you'll need to periodically increase the targetSurface.

You might want to make the target surface a bit bigger than it should be to give the cells some flexibility.

Comment:

content:
Yeah，I have tried this method. In my try, the results is good when the cell is large enough. When the cell is small, tens of lattices long for example, the cell tends to be polyhedron. If I simulate the process that one cell divides several times into hundreds of daughter cells, the cost will be expensive.How to realize that? Thank you~

Q17: Why does field rendering crash CC3D on Mac OS X?

content:
I just installed the 3.7.7 binaries for Mac on a new iMac but with El Capitan.  But for the past two or three Mac binaries, field concentration rendering has crashed CC3D (now) or done nothing.  Cell types run OK.  I am just curious how the other Mac users listed in another post have been able to run it normally.  Thanks

Answer:

content:
Hi Richard

Could you post a toy version of your simulation so that we can take a look at it? Does it happen on ElCapitan only or also on other OSX versions?

Answer:

content:
This is not an answer by the way. I just want to add that I am using CC3D on Mac and also have the same problem. The simulation would crash whenever I switch from Cell to Chemical field, even on a very simple and basic CC3D simulation involving a chemical field. Chemical field seems to work when I draw my own cells using CellDraw rather than the built-in cells. 

Answer:

content:
How can I attach a whole folder of all the .cc3d, .xml, and .py files in the post (in case that might help)? The post only allows me to attach one file at a time when clicking on the '+' symbol.

Answer:

content:
I change the LatticeType from Hexagonal to Square and it worked.

Comment:

content:
simply zip the folder that contains .cc3d, .xml, and .py files and then in the allanswered editor click "+" button (next to "B", "I") buttons and choose "upload image/file"

Q18: control surface and volume simulateously to set up the shape

content:
In my simulation, I try to control the target surface and target volume to set up the shape of the cell. However, the area of the surface is never changed, though the value of lambda is set very high.

attachments(redirected):
.\CC3D_allanswered_linked_files\222.png

Comment:

content:
Could you share more details of the simulation- How are you changing the targetSurface in every mcs and where are you printing it?

Answer:

content:

attachments(redirected):
.\CC3D_allanswered_linked_files\1.png
.\CC3D_allanswered_linked_files\2.png

Comment:

content:
The details are shown in the following pictures. I specify the value of area, "S=pow((3.0*cell.targetVolume)/(4.0*math.pi),2.0/3)*4.0*math.pi ", to guarantee that the cell is sphere during the process. For convenience, there is only one cell at first.

Q19: CompuCell to Model Osmotic Engine

content:
I am a 4th year undergraduate student at Colorado State University studying chemical & biological engineering and biomedical engineering. Currently, I'm in a group working on a class project that requires us to learn a new software package and apply it to a transport phenomena problem. We are working on modeling cancer cells as an osmotic engine in CompuCell. Based on the research we've done so far on the osmotic engine, we are thinking the chemotaxis module would suit our needs. The principle behind the osmotic engine is that a cancer cell can polarize itself in order to move independent of actin and myosin filaments, absorbing water at the leading edge and expelling water out the back, thereby creating movement. Our ultimate goal is to model cancer cell movement as well as volume (and other properties) of cancer cells when exposed to different osmotic environments. Might anyone have any suggestions on how to go about this? We are all very new to using CompuCell, and have been able to create very simple models (i.e. case where water is attracted to cancer cells) but are not sure how to proceed in terms of getting the cancer cells to move in response to the presence of water. We also had emailed Dr. Swat who recommended we look into using compartmentalized cells as well as using toy-mechanism to apply force to cells and thereby simulate movement, depending on the cell's environment. In any case, any advice would be much appreciated.

Thanks,
Elliot

Answer:

content:
How are you planning to model water in this case? If you are considering a `field' like approach, where there are low and high concentrations of water you can probably use one of the diffusion solvers in CC3D. You can specify diffusion rates etc depending on your osmosis. Since the arguments of the field are coordinates in the lattice space, you can use it to specify any form of initial conditions. Now you can attach chemotaxis like attributes to your cell towards this field and you can simulate motion. To measure polarization in response to the chemotaxis you can use the moment of inertia plugin which will give you a major and minor axis indicating how the cell might have changed shape.

Comment:

content:
Yes, we were planning on modeling water as a field. This is probably a really rudimentary question, but is specifying the initial conditions of a field similar to specifying that of a cell? In the xml file, it looks like you can import a txt file that would have the information for the initial concentration field, but I'm not sure what the contents of that txt file would be. Thanks!

Comment:

content:
So the format of the text file is (x_coordinate,y_coordinate,z_coordinate,value). The x,y,z are coordinates between the minimum and maximum positions in your lattice. An example has been provided in the manual. If however you have initial conditions that could be created easily (like don't have to be read from a file) you can specify them in the start function of any class in the python steppable file as well. Something like:

def __init__(self,_simulator,_frequency):
  SteppableBasePy.__init__(self,_simulator,_frequency)
  self.somefield=self.getConcentrationField("field")

def start(self):
  for x in range(0,self.dim.x):
    for y in range(0,self.dim.y):
      self.somefield[x,y,0]=some_value​

Comment:

content:
We have been able to get the basic layout we want setup, however I was wondering how we would get about adding chemotaxis like attributes to the cell. Currently, we are using the chemotaxis plugin, however that is the chemotaxis of water towards our cancer cells, and we would like it to be the other way around. It seems there's not a way to reverse this within the plugin, so we will likely need to make changes elsewhere. Might you have any advice on what plugins/parameters we would need to modify. Thank you very much for all your help.

Comment:

content:
So water has the form of a field and you have cancer cells in the simulation. Does something like this not work? This would be the cancer cell responding to water 'concentration' and moving towards or away from it depending on the sign for value.

def start(self):
  for cell in self.cellListByType(self.CANCERCELL):
    cd = self.chemotaxisPlugin.addChemotaxisData(cell, "WATERFIELD")
    cd.setLambda(value)
    ​

Q20: How about the units in CC3D?

content:
Hi, I wonder what is the unit in a CC3D simulation?  Is it dimensionless? How to make the simulation agree with the experiment?

Answer:

content:
The lattice in CC3D represents a coordinate space in which you can measure positions. or measurements in length (in voxels). The other thing you can visualize from your simulation are measurements in time (in mcs) by observing how your configuration is changing. These length or time values are only a representation . For example, I could have a cell which measures 15 microns in diameter. To represent it I can construct a cell in CC3D with a diameter of 15 voxels and now my scale would be 1 voxel=1 micron. I could also pick a smaller or larger simulation diameter and scale accordingly.

Similarly I could have a problem in which my cell grows, reaches twice its volume and divides. If I know experimentally that the cell division time is 100 minutes, I would from my simulation check how long it takes for its growth and division and relate that mcs value to 100 minutes. 

Comment:

content:
Thank you, I understand that. I still have a question. Can I use the boundary energy to resemble the adhesion between cells? How to make the J(energy) agree with the strength of adhesion i.e. the amount of E-cadherin？

Comment:

content:
So what you can use for this is the adhesion flex plugin (for more details please see the manual). This plugin allows you to specify the adhesion molecule density for each cell (in terms of E-cadherin etc) and writes the resultant surface energy in terms of products of different adhesion molecule concentrations in all the cells. It also allows you to specify the strength of interactions between these surface molecules.

Q21: How to build a circular wall?

content:
Hi, I try to build a circular wall to restrict the scope of cell motility. The codes are as follows. The results are surprisingly interesting. Not only the cell types are different, the keyword, "FREEZE", is also out of use.

attachments(redirected):
.\CC3D_allanswered_linked_files\03.png
.\CC3D_allanswered_linked_files\02.png
.\CC3D_allanswered_linked_files\01.png

Answer:

content:
File attached: image (285.65 KB)

I tried the same code and I have been able to construct the wall with a single cell type. So there could be an issue while you are initializing the other two cell types and something got overwritten. In this code, Freeze seems to work but if you want an alternate way you could set the volume constraint to be very large (cella.lambdaVolume=1000000) and that should essentially do the same thing.

attachments(redirected):
.\CC3D_allanswered_linked_files\image
.\CC3D_allanswered_linked_files\image

Comment:

content:
It seems that the mitosis steppable is the cause. When the volume of the cell satisfies the dividing condition, the interesting phenomenon takes place. Does it mean that the "Freeze" is useless for mitosis? The codes are as follows.

class MitosisSteppable(MitosisSteppableBase):
    def __init__(self,_simulator,_frequency=1):
        MitosisSteppableBase.__init__(self,_simulator, _frequency)
                
        # 0 - parent child position will be randomized between mitosis event
        # negative integer - parent appears on the 'left' of the child
        # positive integer - parent appears on the 'right' of the child
        self.setParentChildPositionFlag(-1)       
    
    def step(self,mcs):
        # print "INSIDE MITOSIS STEPPABLE"
        cells_to_divide=[]
        for cell in self.cellList:
            if cell.volume>500:
                
                cells_to_divide.append(cell)
                
        for cell in cells_to_divide:
            # to change mitosis mode leave one of the below lines uncommented
            self.divideCellRandomOrientation(cell)
            # self.divideCellOrientationVectorBased(cell,1,0,0)                 # this is a valid option
            # self.divideCellAlongMajorAxis(cell)                               # this is a valid option
            # self.divideCellAlongMinorAxis(cell)                               # this is a valid option

    def updateAttributes(self):
        self.parentCell.targetVolume /= 2.0 # reducing parent target volume                 
        self.cloneParent2Child()            

        if self.parentCell.type==self.CONDENSING:
            self.childCell.type=self.NONCONDENSING
        else:
            self.childCell.type=self.CONDENSING​

Comment:

content:
Do you want to divide your wall cell as well? Or is it an error? If you meant only condensing and noncondensing should divide, you can change your line for collecting cell types in the division array to

for cell in self.cellListByType(self.CONDENSING,self.NONCONDENSING):
  if(cell.volume>500):
    cells_to_divide.append(cell) ​

Then only these two types of cells will have updated Attributes. I don't know what kind of output you are getting for the wall cells now but could only be a conflict of current wall target volumes that you are reducing to half.

Comment:

content:
Yeah, thank you so much. It is an error since I misunderstand the function of "Freeze". 

Q22: How to simulate an elastic ring in 2D or an elastic hollow sphere in 3D?

content:
I want to simulate an elastic ring in 2D or an elastic hollow sphere in 3D. There are cells dividing in the ring/sphere. When the ring is full of cells, it will exert a contraction force on the cells. How to achieve? Thank you!

Answer:

content:
I would use FocalPointPlasticity Plugin (FPP).  FPP will connect centers of mass of nearby cells with elastic link. However when your cells are dividing then you woudl need to manually break the links and reestablish then after the division happens. The reason for that is that you want links for form a "chain" where one segment of the "chain" connects nearby cells. What you do not want is to have a chain where a segment of a chain "skips" new cell that appeared as a result of the mitosis
.

Answer:

content:
To actually generate the circle or sphere check out the code snipets and CC3D project file at https://www.allanswered.com/post/pgrnm/i-am-trying-to-mimic-the-segment-patterning-in-fruit-flies-through-a-cc3d-model-can-anyone-help-me-as-to-how-to-draw-out-a-3d-paraboloid-piff-file-with-different-cell-types-along-the-surface/

You would then add focal point plasticities to keep the circle/sphere layout.

Comment:

content:
Actually, the background is that a non-living elastic ring exerts a tension on the cells inside the ring. I want to study the influence of the elastic ring for the division of cells. Besides, I am trying to mimic the phenomenon when the ring is snipped.  When I use the FPP plugin, the cells constituting the ring are easy to separate from each other. How to avoid it?

Comment:

content:
Actually, the background is that a non-living elastic ring exerts a tension on the cells inside the ring. I want to study the influence of the elastic ring for the division of cells. Besides, I am trying to mimic the phenomenon when the ring is snipped.  When I use the FPP plugin, the cells constituting the ring are easy to separate from each other. How to avoid it?

Q23: VTK output problem when running AdhesionFlexPython

content:
Dear Admins,

I tried running the AdhesionFlexPython demo on both CC3D 3.7.7 and 3.7.6 in Mac OSX High Sierra. I encountered a strange problem where the moment the adhesions are changed by the AdhesionFlexSteppables.py script, the VTK outputs all zeroes for all entries (such as CellID, CellType, ClusterID).

I verified this by changing the original script to

# Same as demo until this point....

if mcs==100:
            for cell in self.cellList:
                print "CELL ID=",cell.id, " CELL TYPE=",cell.type
                if cell.type==1:
                    # accessing adhesion molecule density using its name
                    print "NCad=", self.adhesionFlexPlugin.getAdhesionMoleculeDensity(cell,"NCad")

                    # accessing adhesion molecule density using its index -
                    # molecules are indexed in the sdame order they are listed in the xml file
                    print "Int=", self.adhesionFlexPlugin.getAdhesionMoleculeDensityByIndex(cell,1)

# And so on....​

# Same as demo until this point....

if mcs==101:
            for cell in self.cellList:
                print "CELL ID=",cell.id, " CELL TYPE=",cell.type
                if cell.type==1:
                    # accessing adhesion molecule density using its name
                    print "NCad=", self.adhesionFlexPlugin.getAdhesionMoleculeDensity(cell,"NCad")

                    # accessing adhesion molecule density using its index -
                    # molecules are indexed in the sdame order they are listed in the xml file
                    print "Int=", self.adhesionFlexPlugin.getAdhesionMoleculeDensityByIndex(cell,1)

# And so on....​​

runScript.command -i $working_dir/$working_script -f 100 --currentDir $working_dir -o $working_dir/movies --exitWhenDone​

Here's a video of the default run, I don't understand what's happening here, why are all the cells becoming type 0 (medium)?

File attached: AdhesionFlex.mpg (2.61 MB)

attachments(redirected):
.\CC3D_allanswered_linked_files\AdhesionFlex.mpg

Comment:

content:
Did anyone else see this problem or it's only an issue with my simulations? Any response is greatly appreciated so that I can troubleshoot the problem.

Answer:

content:
I ran the AdhesionFlex Simulation with the VTK output but could not reproduce the problem. Could you post the exact simulation that you ran?

Comment:

content:
I ran the default AdhesionFlexPython demo on both Mac OSX High Sierra and Ubuntu 14.04 with CC3D 3.7.7. The CC3D on Ubuntu was compiled by myself. Both these cases showed the same result of Type 1 disappearing slowly.

I have attached the .zip file of the demo that I ran just to be sure:
https://mega.nz/#!5AoCSDwB!Jw9SXf2yDz3wV0hNIQKInEtLqN_yGnCa6NvJS7v0rw8

Answer:

content:
Any ideas what might be the issue or any ways of troubleshooting? Here's some sample outputs from running:

File attached: LatticeData.zip (459.2 KB)

attachments(redirected):
.\CC3D_allanswered_linked_files\LatticeData.zip

Answer:

content:
After tweaking things about, the only way that I can get the simulation to have a logical VTK output is if the medium has no adhesion molecules whatsoever. I have attached the sample scripts and outputs here:
File attached: output.zip (3.55 MB)

Any comments about why this is so will be greatly appreciated.

attachments(redirected):
.\CC3D_allanswered_linked_files\output.zip

Q24: I am trying to mimic the segment patterning in fruit flies through a CC3D model. Can anyone help me as to how to draw out a 3D paraboloid PIFF file with different cell types along the surface?

content:
Currently, I see no options to draw a 3D PIF in Cell Draw. The objective is to show segment width of different cell types (coloured differently) with variable values of the contractility term aka Young's modulus. Also, the environment needs to be bounded, meaning the boundary conditions should be fixed or periodic but can't be open.

Comment:

content:
Hi, as far as I am concerned, the contractility term is a little different from Young's modulus. What's your contractility term, the lambda volume or surface? 

Comment:

content:
Not very familiar with Cell Draw. But you can use a separate CC3D simulation to generate a PIF file with your initial conditions first and later on export it for other simulations.  (CC3D Gui -> Tools-> Generate a PIF from current snapshot) . Advantage is you can attach your Young's modulus to each cell and adjust it's constraints to describe your system. (Note to construct cells of arbitrary lengths you can use self.cellField)  

Comment:

content:
The contractility term is Lambda surface

Comment:

content:
Yes, when I use the Lambda surface, the cells often tend to be rectangle or so, which is not analogous to the condition of cells. Have you ever meet such things?

Comment:

content:
Yes, in a realistic embryo system of a developing fly the cell shapes are more of polygonal nature with pentagons and hexagons visible but with CC3D if I initialize the system with square cells, ideally the cells will try to get to a circular shape over time due to energy minimisation principle. The catch point is the regional geometrical constraints play a role (i.e. if you dont want your cells to divide and play with a constant number of cells).

Although I must say, even though visibly they look rectangular or squarish from the sim output, performing a Voronoi analysis (after  extracting the center of mass of the cells) showed a distribution of polygonal shapes at steady state!

Answer:

content:

attachments(redirected):
.\CC3D_allanswered_linked_files\SphereOfCells.zip
.\CC3D_allanswered_linked_files\sphereofcells.png

Q25: if I changed some parameter during the simulation running, will the result be affected?

content:

Answer:

content:
In my experience, no. If you stop the simulation and restart, the parameters changed in the XML will work. Why don't you have a try? Then you can confirm it.

Answer:

content:
Yes. If you change parameter in the Model Editor window it does affect the running simulation. For instance if you change temperature in the model editor new value of the temperature will be used in subsequent MCS.

Model Editor-

attachments(redirected):
.\CC3D_allanswered_linked_files\Screen Shot 2017-11-27 at 12.29.45 PM.png

Q26: How to control the shape of a cluster of cells?

content:
Hi, everyone. In my simulation, there is a cluster of cells undergoing cell division. I wonder if I can get the  perimeter L of them and add an energy term,  $\frac{1}{2}k\left(L-L_0\right)^2$12 k(L−L₀)² .
I want to simulate the intercellular actin cable et al. in this way.

Answer:

content:
That looks like the surface area plugin with   $\lambda_{surface}=\frac{k}{2}$λ_surƒ ace=k2 . (In 2d, it generalizes to a perimeter. Should help control `pixel' distribution at the periphery)

Comment:

content:
Can the plugin be used for a cluster of cells (i.e. control the total surface area of these cells not for each cell)?

Comment:

content:
I might have misunderstood your original question. By `perimeter' you are referring to the resulting shape at the boundary the cluster is forming (are your cells in a ring?) I don't believe you should directly try adding a term that signifies the target length arising from a group of single cells. Best case scenario would probably be using focal point plasticity links(similar to your equation form) between neighboring cell pairs in the cluster in order to control L0,like breaking up L0 into smaller pieces.    

Comment:

content:
Actually， the shape is a disc or other polygons, not only a ring. I want to apply the constraint, $\frac{1}{2}k\left(L-L_0\right)^2$12 k(L−L₀)² , on the outermost ring of cells. The biological background is, the outermost cells contract to exert a force on the inners. Both the outer and inner cells are of the same type and they are under mitosis. Is it possible to simulate this in CC3D?
Besides, what's the meaning of volume and surface in 2D simulation? I do not see the description in the CC3D tutorial. For example, a cell is made up of 2*2 pixels in 2D, then the surface is 8 and the volume is 4? 
Thank you.

Comment:

content:
So the point I was trying to make was you can only break your constraint length into smaller lengths and control it using FPP. The Hamiltonian in CC3D is a sum over individual cell components and their interactions. 

In 2d, cell volume would generalize to the total number of pixels for the cell. So if you have a 'squarish' cell of 3 by 3, the total number of pixels in this case is 9 which is the volume. Cell surface is given by how the cell is in contact with the boundary, in this case 3*4=12.

Comment:

content:
Yeah， I will have a try. Thank you so much.

Comment:

content:
Hello, Yuan

 My name is Jingwei Qiu, I am an artist now basic between Beijing and SFO. I am working on a project that need somebody familiar this model , but is hard to find programer who have the knowledge both in bio-cell and computing, so let me know can I ask the help from you, we can discuss the treatment。 Thank you

Comment:

content:
Hello Jingwei,
I am not sure whether I can meet your requirements. So what do you want to do specifically?

Comment:

content:
Thank you so much for your reply! can we add a qq or wechat? So that I can share more detail with you.
my QQ:1071787728 Thank you!

Comment:

content:
This program is barely used by people,its very professional.All my programmer friend can only understand few function in it.

Comment:

content:
Hey, Yuan

 Do you normally use QQ? I would love to go further in that topic with you,Let me know how you think)

Q27: Would SteadyStateDiffusionSolver2D be appropriate to be used with Chemotaxis

content:
I have my model based on the paper "Emergent Stratification in Solid Tumors Selects for Reduced Cohesion of Tumor Cells: A Multi-Cell, Virtual-Tissue Model of Tumor Evolution Using CompuCell3D" for tumor growth. However, I would like to have my cells to move  towards higher glucose and additionally oxygen. Both these nutrients would have high diffusion constants. Using "DiffusionSolverFE" is making my simulations very slow as has been mentioned in the tutorials. Hence, I was wondering if I use SteadyStateDiffusionSolver  as has been used in the above paper, would chemotaxis be appropriate? 

Comment:

content:
You should be able to use any module for solving your field and use that for chemotaxis. When you change the diffusion solver, also update the ChemicalField Source under the Chemotaxis plugin. 

Answer:

content:
Any solver can be used with chemotaxis but for large diffusion constants we typically use SteadyState solver that essentially solves Helmholtz equation in Cartesian coordinates

(d/dx)(du/dx) + (d/dy)(du/dy) + lambda*u = f(x,y).

As you can see there is no dependence on time here and this is the limiting case for the diffusion equation when d/dt term is zero because the field stopped evolving (i.e. at large t.) Clearly this is an approximation so you have to be aware of it and double check if this approximation is appropriate for your problem. However using the field from SteadyState solvers in Chemotaxis plugin is fine

Q28: fail to generate PIF file from current snapshot

content:
Hi, everyone, I am trying to generate the PIF file and use it in the following simulation. I get a "fail to respond " error, just as the picture.

attachments(redirected):
.\CC3D_allanswered_linked_files\123.png

Comment:

content:
So the screen becomes unresponsive once you click on 'Generate PIF file from current snapshot'? I used to have an issue when the command can't be selected at all. The way I solved it is instead of running the simulation by hitting play initially, I would hit step on the first mcs and then click on play later (Does that help in any way? ). That seemed to work with the command being available after that. Though this problem is fixed in the newer cc3d versions. 

Comment:

content:
Yeah, once I click on  'Generate PIF file from current snapshot' it becomes unresponsive. If I hit the step on only and then click to generate PIF, the same result is got.

Comment:

content:
There is also an option for PIF Dumper Steppable (Its in the manual) which you can add to the main python file. This way you don't have to use the player. Does that work for you?

Comment:

content:
Yes， this is an effective way. Thanks sincerely.

Answer:

content:
Does the solution work?

Comment:

content:
Yes, I succeed to generate the PIF using the PIF Dumper Steppable.

Q29: Can I simulate the lumen formation in tissues?

content:
Hi everyone, I try to simulate the formation and fusion of lumens in the tissue. The sketch is as follows. The black hole is the lumen. Each lumen is between two cells.


I wonder if you could give me some advice, how can I simulate the initial lumen in CC3D? In my scenario, the lumen may exert a pressure on the surrounding cells so that it can extend. Besides the cells need to contract by themselves. Thank you!

attachments(redirected):
.\CC3D_allanswered_linked_files\lumen.png

Answer:

content:
I am curious about this as well. What type of tissue is this? If in 2D, you could initialize a cell type for the lumen "space" 

Answer:

content:
Yes, the best way to model a lumen is probably as "cells". Basically, blobs of water (or blood, or ...)

Answer:

content:
See this paper for details https://www.ncbi.nlm.nih.gov/pubmed/27193300

Comment:

content:
The paper is very detailed and a good source- thank you for the suggestion! We need a Journal Club on the forum!

Q30: An error happens when I try to open the model in player

content:
File attached: model.rar (18.01 KB)

attachments(redirected):
.\CC3D_allanswered_linked_files\QQ截图20171208095803.png
.\CC3D_allanswered_linked_files\QQ截图20171208102236.png
.\CC3D_allanswered_linked_files\model.rar

Answer:

content:
could you post the example you are trying to run. i n general I would stick to newer versions of CC3D as they ship with complete Python distribution. BTW the above error happens because you have not imported random  in Python. If you type

import random

random.rand()​

it should work

However , depending on your simulation you may want to explore numpy's random  utilities in that cas you woudl type

import numpy.random as nr

# now you can type

nr.rand(3,2)

# and this is what you will get
# array([[ 0.14022471,  0.96360618],  #random
#       [ 0.37601032,  0.25528411],  #random
#       [ 0.49313049,  0.94909878]]) #random

see https://docs.scipy.org/doc/numpy-1.13.0/reference/generated/numpy.random.rand.html#numpy.random.rand

Comment:

content:
Thank you. The model is attached now. I don't even use the function in Numpy. So I think the error may due to other reasons. BTW is there a no gpu version for 3.7.7 or 3.7.6 ?

Comment:

content:

In addition to that, there are still two problems for 3.7.7.

1. It seems that there is no CellDraw in 3.7.7
2. If I click the Compucell3D immediately, it will crash in splash screen.

Answer:

content:
Thanks for reporting the startup crash issue. We have fixed it.And you are right there is no cell draw in 3.7.7. 

Answer:

content:
The issue with your code was not the model itself but the cc3d file

This is how it looked like:
<Simulation version="3.6.2">
   <XMLScript Type="XMLScript">Simulation/lumen.xml</XMLScript>
   <PythonScript Type="PythonScript">Simulation/lumen.py</PythonScript>
   <Resource Type="Python">Simulation/lumenSteppables.py</Resource>
   <PIFFile>C:\CompuCell3D-64bit\(u'D:\cc3d\lumen\lumen.piff', u'All Files (*)')</PIFFile>
</Simulation>


and this is how it should look like:

<Simulation version="3.6.2">
   <XMLScript Type="XMLScript">Simulation/lumen.xml</XMLScript>
   <PythonScript Type="PythonScript">Simulation/lumen.py</PythonScript>
   <Resource Type="Python">Simulation/lumenSteppables.py</Resource>
</Simulation>

Simply you have specified a path of the PIFFile that did not point to anything and thi sis the reason you have seen the crash

If you want to specify PIFFile path in the .cc3d file you could do the following:

<Simulation version="3.6.2">
   <XMLScript Type="XMLScript">Simulation/lumen.xml</XMLScript>
   <PythonScript Type="PythonScript">Simulation/lumen.py</PythonScript>
   <Resource Type="Python">Simulation/lumenSteppables.py</Resource>
   <PIFFile>Simulation/lumen5.piff</PIFFile>
</Simulation>


File attached: model_yuan_fixed.zip (20.91 KB)

Let us know if this simulation runs on your machine. I have  tested it on windows with 3.7.7. binary

attachments(redirected):
.\CC3D_allanswered_linked_files\model_yuan_fixed.zip

Comment:

content:
So do I have to install an older version if I want to use CellDraw?

Answer:

content:
For the time being, yes. How complicated is the geometry you are creating? 

Comment:

content:
Yes, the model runs normally now. Thanks sincerely.

Comment:

content:
Just as the following picture:

attachments(redirected):
.\CC3D_allanswered_linked_files\QQ截图20171211100024.png

Q31: How can I obtain the velocity of a cell during the simulation?

content:
Hello everyone. I simulate the migration of a cell over a substrate and I would like to obtain the velocity of this cell and the time of each iteration. Do you know if there is a way to know them?
Thank you!

Answer:

content:
You can find a velocity measure by tracking positions of center of masses of cells and the time taken to attain the difference. So if the center of mass of the cell has shifted from x1 to x2 in t mcs, the velocity would be (x2-x1)/t

Answer:

content:
One thing to remember though is to be aware that is t=1 MCS (Monte Carlo Step) you will get a very noisy estimate of the velocity so it is best to use longer time periods. the optimal interval to use may depend of the actual simulation you run so you may need to play a bit with different values

Answer:

content:
Thank you for the answer. Do you know where I can find data such as the position of the center of mass of the cell during the simulation (because I try to use the plugin "CenterOfMass" but I can't find the data for post-processing)?

Answer:

content:
If you use twedit++ there is a code snipet at <CC3D Python><Cell Attributes><Center of Mass>, which inserts the code below.
# Make sure CenterOfMass plugin is loaded
# READ ONLY ACCESS
xCOM=cell.xCOM
yCOM=cell.yCOM
zCOM=cell.zCOM

Answer:

content:
This is a little embarrassing, but I have used a ruler on screenshots taken at simulation start and end points to measure # of pixels/voxels traveled by a clump of cells to get their tissue-invasive velocity. The simulation was of several hours' movement, so this was basically a long, running average, as mentioned above. Whether I did it on the screen or printed on paper I can't remember, but either is fine -- though you have a record if you keep the paper. The traveling cells stayed visible in one plane of a 3-D sim, coincidentally. Not to discourage tracking it computationally -- you could just as easily get a start and end point and do the same calculation -- just another way it has been done.

Q32: A question from the paper.

content:
Hi, the paper "Multi-Scale Modeling of Tissues Using CompuCell3D" says, "The fixed discretization makes explicit modeling of fibers or membranes expensive, since the lattice constant must be set to the smallest scale to be explicitly represented. Cell membrane fluctuations are also caricatured as a result of the fixed spatial resolution. However, the latest versions of CC3D support a layer of finite-element links which have length but zero diameter. These can be used to represent fibers or membranes, allowing a simulation to combine the advantages of both methods at the cost of increased model complexity."
How to realize it?

Answer:

content:
The layer fo finite-element-like links is essentially Focal Point Plasticity. It places a constraint on the distance between two centers of masses of the two cells. For example you can have a compartmentalized cell composed of many compartments and you can connect "outermost" compartments with a series of focal point plasticity links. This could provide contractile force, of  similar type as membrane would exert. 

Q33: what's the meaning of neighborAddress?

content:
Hi, everyone. I am just reading the CC3D code in Ph.D. thesis 'Modeling interactions between a tumor cell and its host epithelium', Eline Boghaert, Princeton University (2014).
There are some puzzles for me in the following code, what're the properties of  neighborAddress ?
 diameters?

 

#for each LEP calculate the number of cells that are 2.5 cell diameters away (2nd order neighbors)
class ApoptosisSteppable(SteppableBasePy):
    def __init__(self,_simulator,_frequency=1):
        SteppableBasePy.__init__(self,_simulator,_frequency)
    def step(self,mcs):
        if mcs>10:
            cell_apoptosis=[]
            for cell in self.cellListByType(1):
                neighbors=[]
                a=cell.id
                id=[cell.id]
                cellNeighborList=self.getCellNeighbors(cell)
                for neighbor in cellNeighborList:
                    if neighbor.neighborAddress and neighbor.neighborAddress.type==1:
                        neighbors.append(neighbor.neighborAddress)
                        id.append(neighbor.neighborAddress.id)
                        
                for cell in neighbors:
                    cellNeighborList=self.getCellNeighbors(cell)
                    for neighbor in cellNeighborList:
                        if neighbor.neighborAddress and neighbor.neighborAddress.type==1:
                            if neighbor.neighborAddress.id in id:
                                pass
                            else:
                                id.append(neighbor.neighborAddress.id)
            #if there are more than 10 cells within a 2.5 cell diameter radius the LEP undergoes apoptosis with a given probability
            #the randint function is used to specify this probability
                if len(id)>10 and randint(1,1000)<=10:
                    cell_apoptosis.append(a)
            
            for cell in self.cellListByType(1):
                if cell.id in cell_apoptosis:
                    cell.targetVolume=0
                    cell.lambdaVolume=1000
                ​

Answer:

content:
strictly speaking neighborAddress is a C++ address of the neighbor cell object. In Python neighborAddress is simply a neighbor cell so if you type:

neighborAddress.id
you will get id of th neighbor cell

if you type
neighborAddress.volume

you will get volume of the neighbor cell

etc..


One thing to keep in mind is that Medium can be a neighbor too and Medium is a zero pointer in C++ so the neighbor address of it will be 0 and you cannot as for id or volume of the Medium. This is why you often see the if statement that checks if neighbor address is NULL or not

Also , the new examples in 3.7.6 + use somewhat cleaner terminology w.r.t. iteration over neghbors

  

Comment:

content:
Oh，I see, thank you. I still have a puzzle, when I meet some commands which don't appear in the manuals, how can I know its usage and function, in addition to asking in the community?

Answer:

content:
I found the new Python Scripting Manual helpful in this particular case http://pythonscriptingmanual.readthedocs.io/en/latest/SteppableBasePy_class.html .
Most of the simulation is done on self.cellList, and I was able to understand the Python scripting more once I knew more about the basic class. 

I like the particular example you posted. I am also doing simulations involving logic based on cell neighbors. 

Comment:

content:
Usually the best thing is to ask people in this forum. We are working to fix the documentation but sometimes  code development cycle is ahead of the documentation cycle. Any questions you post here will get incorporated into the manuals. It also help us understand what aspects of CC3D are most confusing to people 

Q34: Conversion factors

content:
For the paper "Emergent Stratification in Solid tumors ..." there is a conversion factor from fmol/cell to fmol/voxel (0.1fmol/cell/s to 2.25fmol/voxel/MCS). I can get the conversion from MCS to sec. However, what is the factor used for conversion of  per cell to per voxel. The paper mentions using a scaling factor to convert 5mM to 0.32fmol/voxel. 

Similar conversion factors are also used in "Multiscale Modeling of Tissues " , the book chapter. 

I would highly appreciate any help. 

Answer:

content:
I think in this simulation we used steady state diffusion solver that for a given initial and boundary condition outputs concentration numbers for each pixel. All we are saying in the paper is that whatever the solver outputs will be interpreted in such a way that when the solver outputs 0.32 fmol we interpret it as 5mM of real concentration. Simply put I can rescale concentration in any way I want i.e. when i keep diffusion constant fixed and rescale initial concentration by a factor F the field will evolve in the same way as the field that was not scaled. the only difference will be scale. 

Comment:

content:
For the same paper would you be able to explain how the secretion rate of glucose from stromal cells was calculated in  the absence of the tumor. And also the threshold of 102.0 was assumed based on the cell being exposed to zero glucose concentration for 24 hours (240 MCS). I would highly appreciate if you could kindly shed some light on the calculation steps to reach that value. Thanks so much.

Answer:

content:
given we set voxel area to 16 um^2, voxel volume will be 64um^3:

5mM= 5e-3 mole * (1/L) * (64um^3/voxel) = 0.32e-15 mole/voxel

0.1 fmol/cell/s = 0.1 fmol * (1/cell) * (cell/(16 voxels)) * (1/s) * (360 s/MCS)=2.25 fmol/voxel/mcs

Q35: Fail to run the model from CC3D web

content:
I have just run the model code of the PKD paper( http://www.compucell3d.org/Models/PKD).
There seem to be some problems,
1. when I click the 'Open in Player', the CC3D player doesn't work (both of the isoCyst and tubule model).
2. it seems that the  Graphs.py is lost in the tubule folder,
<version 3.7.7, Windows 7, 64-bit.>

<Simulation version="3.7.3">
    <PythonScript Type="PythonScript">PKD.py</PythonScript>
    <Resource Type="Python">PKDSteppables.py</Resource>
    <Resource Type="Python">Graphs.py</Resource>
</Simulation>​

Answer:

content:
We fixed it

here is the link

http://www.compucell3d.org/Models/PKD

If it does not work I am attaching it here as well

File attached: PKD_paper_codes_fixed.zip (40.93 KB)

attachments(redirected):
.\CC3D_allanswered_linked_files\PKD_paper_codes_fixed.zip

Q36: Python keeps crashing and compucell3D fails to open.

content:
Hi All,

I have been using compucell3D for over a year without a problem, but Since downloading the latest version, Compucell3D player fails to open unless via tweddit, and then python keeps crashing at random times during the simulation. This is not an error in the code I have checked this for sure. But even when I uninstall the latest version and reinstall an older version of compucell3D I am having this problem. has anyone else had this issue or knows what I could do.

Thanks,

Daniel

Answer:

content:
If this happens on Windows you may need to reinstall CC3D. Preferably you may need to manually remove the entire CC3D installation directory prior to installation. Let us know if this continues to be the case

Comment:

content:
Thank you, I did this and reinstalled version 3.7.5 and it is now working I shall try again with version 3.7.7 later.

Comment:

content:
I have the similar problem with the CC3D player for 3.7.7 and 3.7.6, so I just open it via twedit. The 3.7.5 no GPU version is normally running for me.

Answer:

content:
See the solution above -  manually removing the CC3D installation folder on Windows Prior to installation of C3D 3.7.7 will solve the problem

Comment:

content:
It seems useless for the CC3D player, however, open it via twedit is okay.

Q37: The execution sequence of the plugins and steppables

content:
The following is the pseudo-code of CC3D. My question is what is the execution sequences of the plugins and steppables?   Are the sequences impacting on the results? What should I do if I want to make the steppables run in my sequence?

attachments(redirected):
.\CC3D_allanswered_linked_files\QQ截图20171215170902.png

Answer:

content:
One way to enforce a particular ordering of steppables would be to simply include them all in a single steppable. You then have complete flow control in python.

Comment:

content:
Yeah, a good idea! Thank you so much！

Answer:

content:
Steppables are executed in the same order in which they are aded to steppableRegistry. So in your main python file you typically see the following

from scientificHistBarPlotsSteppables import HistPlotSteppable
histPlotSteppable=HistPlotSteppable(_simulator=sim,_frequency=1)
steppableRegistry.registerSteppable(histPlotSteppable)

from scientificHistBarPlotsSteppables import BarPlotSteppable
barPlotSteppable=BarPlotSteppable(_simulator=sim,_frequency=1)
steppableRegistry.registerSteppable(barPlotSteppable)

CompuCellSetup.mainLoop(sim,simthread,steppableRegistry)​

Comment:

content:
BTW, what about the order of plugins?

Q38: Disappearing cells

content:
Dear reader,

I loose the record of some cells at the end of my simulation, what could be the cause for it?

An example file extraction:

cell_no,time_mcs,x_position_(px),y_position_(px)
...
428,966,22.6896551724,14.8965517241
429,966,22.56,36.12
428,967,22.6153846154,15.1923076923
429,967,22.7083333333,35.9166666667
428,968,22.7037037037,14.851851

STOPS?? :)

There should be 999 entries? What to do/how to fix? 

Best regards,
Pauli
:)

Answer:

content:
Could you post the part of your code that is printing to the file? A possible reason could be if you are not `flushing' after printing (eg print>>some_fileHandle, results ; some_fileHandle.flush() ) but this is just a guess.  

Answer:

content:
Are you sure the missing lines aren't actually in the file somewhere other than at the end? Depending on your data structure Python might not be printing things in the order you expect. Dictionaries in particular can print with odd ordering.

Answer:

content:
Hi,

Thank you for your comments!
Here is the steppables.py file almost as it is:

from PySteppables import *
import CompuCell
import sys
import random
from math import *
from XMLUtils import dictionaryToMapStrStr as d2mss
from PySteppablesExamples import MitosisSteppableBase

class ConstraintInitializerSteppable(SteppableBasePy):
def __init__(self,_simulator,_frequency=1):
SteppableBasePy.__init__(self,_simulator,_frequency)
def start(self):
for cell in self.cellList:
cell.targetVolume=30
cell.lambdaVolume=1.0

class peittoSteppable(SteppableBasePy):

def __init__(self,_simulator,_frequency=1):
SteppableBasePy.__init__(self,_simulator,_frequency)
self.cellA=None
self.cellB=None
def start(self):

self.pW = self.addNewPlotWindow(_title='Tracking', _xAxisTitle='Y',
_yAxisTitle='Variables', _xScaleType='linear', _yScaleType='linear')

for cell in self.cellListByType(self.NPCELLS):
c = random.choice(['green','yellow','red','blue']) #,'yellow','red','blue'
self.pW.addPlot(str(cell.id), _style='Lines', _color='green', _size=5)

for cell in self.cellListByType(self.ACELLS):
c = random.choice(['green','yellow','red','blue']) #,'yellow','red','blue'
self.pW.addPlot(str(cell.id), _style='Lines', _color='blue', _size=5)

FileName = "C:/CompuCell3D-64bit/pauli/movements/m16/okok/heihou99.txt"
FileName2 = "C:/CompuCell3D-64bit/pauli/movements/m16/okok/tt2.txt"

self.File = open(FileName,"w")
self.File2 = open(FileName2,"w")

self.File.write("cell_no, time_mcs,x_position_(px),y_position_(px)\n")
self.File2.write("cell_no, distances, time_mcs,x_position_(px),y_position_(px)\n")

def step(self,mcs):
for cell in self.cellList:
self.File2.write(str(cell.id)+","+str(mcs)+","+str(self.distanceBetweenCells(self.attemptFetchingCellById(15),self.attemptFetchingCellById(cell.id)))+","+str(cell.xCOM)+","+str(cell.yCOM)+"\n")

for cell in self.cellListByType(self.NPCELLS):
self.pW.addDataPoint(str(cell.id), cell.xCOM, cell.yCOM)
self.File.write(str(cell.id)+","+str(mcs)+","+str(cell.xCOM)+","+str(cell.yCOM)+"\n")
for cell in self.cellListByType(self.ACELLS):
self.pW.addDataPoint(str(cell.id), cell.xCOM, cell.yCOM)
self.File.write(str(cell.id)+","+str(mcs)+","+str(cell.xCOM)+","+str(cell.yCOM)+"\n")

def finish(self):
# Finish Function gets called after the last MCS
pass

-Pauli
:)

Answer:

content:
Try self.File.flush() and self.File2.flush() after the write statements. I tried a version of your code and it works for me

Answer:

content:
Instead of the the flush commands you might just need to explicitly close your output files. Closing will flush the write buffer.  I believe this will do it;

def finish(self):
    # Finish Function gets called after the last MCS
    self.File.close()
    self.File2.close()

Answer:

content:
Another way to flush the writes is to open your output file(s) with a zero size buffer;
self.File=open(FileName,"w",0)
The zero is the output buffer size. This flushes the output after every write command.

If you don't write to the file too often then the zero size output buffer wont slow things down.

Answer:

content:
Thank you for your help!
I see that these code lines should solve the issue. I test tomorrow and report if still any related problems.

Comment:

content:
Thank you for your help!
I see that these code lines, yours and James', should solve the issue. I test tomorrow and report if still any related problems.

Comment:

content:
thank you for the code, should be ok

Q39: Some questions about the commands

content:
Dear all， I have some questions on the following codes from the PKD paper.

    def dePixelating(self,cell):
        L=[];  compList=self.inventory.getClusterCells(cell.clusterId)
        for cell2 in compList:
            if (cell2.id != cell.id):
                pixelList=self.getCellPixelList(cell)
                for pixel in pixelList:
                    L.append(pixel.pixel)
        for pixel in L:
            self.cellField.set(pixel,cell)
        cell.targetVolume+=len(L)​

1. what's the function of self.cellField.set(pixel,cell)? To make the pixel belong to the cell?
2. should the command "pixelList=self.getCellPixelList(cell)" be "pixelList=self.getCellPixelList(cell2)"?
3.what 's the function of this function? 

Answer:

content:
As I understand it, this is a function under the Mitosis Steppable Class, for the Tubule Python code. Cell2 is an index going through the compList.  The function is called using (self, cell), so I assume we are interested in the pixels of cell and not cell2.

I think this is related to mitosis or dividing of clustered cells, as opposed to non-clustered cells. See pg 79-81 of the 3.7.2 Reference Manual. The general 3 steps to that includes getting the set of all pixels contained by the clustered cell. Then pixels are assigned to either the parent or child cell.

Q40: SteadyStateDiffusionSolver2D Equation

content:
Hello everyone!

I am working on a project using CompuCell3D, and I am using the PDE Solver "SteadyStateDiffusionSolver2D" to manage the diffusion processes of Oxygen.

However, I can't make sense of the results in the simulation. I have tried to solve the equation given in the documentation (Helmholtz equation) numerically, using Matlab, and the solution is very different.
For instance, in CompuCell3D, there are negative values, which I find unreasonable since I have no cells consuming the substance (there is a single cell diffusing oxygen in the centre of the simulation lattice).
Moreover, I couldn't find any information online regarding the use of the Helmholtz equation (in the form provided in the documentation) for diffusion processes.

Could you please clarify me about this issue or provide me with any documentation that details the methods used in this solver?

Thank you for your time!

Comment:

content:
This from the manual: Moreover the secretion constant needs to have negative value if we want to secrete positive amount of substance - this weird requirements comes from the fact that we re using 3 rd party solver which inverts signs of the secretion constants.

Does inverting the sign for the secretion solve the problem for negative values?

Comment:

content:
Thank you for your suggestion! However, I believe that condition does not apply to my case. I am managing secretion in the XML file and, from what I understood from the Reference Manual, that specification refers only to secretion controlled in Python. In fact, the "SteadyStateDiffusionSolver" Demo that manages secretion in the XML file also has positive values for secretory cells.

Answer:

content:
Could you post a  simplified version of your simulation?

Answer:

content:
This is the code I am using in my simplified version of the simulation (with only one cell in the centre of the lattice):

<Potts>
  <Dimensions x="500" y="300" z="1"/>
  <Steps>1000</Steps>
  <Temperature>10.0</Temperature>
  <NeighborOrder>2</NeighborOrder>
</Potts>

<Plugin Name="CellType">
  <CellType Freeze="" TypeId="0" TypeName="Medium"/>
  <CellType Freeze="" TypeId="1" TypeName="Produtor"/>
</Plugin>

<Steppable Type="SteadyStateDiffusionSolver2D">
  <!-- Specification of PDE solvers -->
  <DiffusionField Name="O2">
     <DiffusionData>
        <FieldName>O2</FieldName>
        <DiffusionConstant>2500</DiffusionConstant>
        <DecayConstant>0</DecayConstant>
     </DiffusionData>
     <SecretionData>
        <Secretion Type="Produtor">1</Secretion>
     </SecretionData>
     <BoundaryConditions>
        <Plane Axis="Y">
           <ConstantDerivative PlanePosition="Max" Value="0.0"/>
           <ConstantDerivative PlanePosition="Min" Value="0.0"/>
        </Plane>
        <Plane Axis="X">
           <Periodic/>
        </Plane>
     </BoundaryConditions>
  </DiffusionField>
</Steppable>​

(Besides this portion, there's the PIF Initializer)

I am running tests with the Decay Constant as 0 or as 0.39.
When it is is 0, this is the result: Max = 0.0260 and Min = -0.0038

On the other hand, in the Matlab simulation (solving the equation with finite differences and applying Jacobi Method), this is the result: Max = 0.0258 and Min = 0

There are no negative values, although there is a small hole in the centre.

In the simulation with the Decay Constant as 0.39, both graphs have only positive values, but the shapes follow the same pattern as above (peak in CC3D vs cone in Matlab simulation) and the maximum values still differ.

My goal is to understand how the values entered in the XML file evolve during the simulation so that I can determine a realistic set of constants for my actual project.

Thank you!

attachments(redirected):
.\CC3D_allanswered_linked_files\Plot-2500-0-1.png
.\CC3D_allanswered_linked_files\Plot-2500-0-1.png

Answer:

content:
When you run Matlab code are you solving Helmholtz equation or are you solving the diffusion equation and run it for a long time to determine the steady state of the solution?

File attached: diffusion_steady_state_ext_potential.zip (7.37 KB)

I am attaching mine example that is very similar to yours. I noticed that that when decay constant is 0 the solver does get unstable. From what I remember with steady state solvers the decay constant had to be different than 0.0. So when I used decay constant of 0.4 the solution looks reasonable.

As a matter of fact the comments in the code for steady state solver state that when decay constant is 0.0 the solution may not exist:

/* elmbda */
/* the constant lambda in the helmholtz equation. if */
/* lambda .gt. 0, a solution may not exist. however, hwscrt will */
/* attempt to find a solution. */


I am also attaching original code of th steady state solver for 2D case. Take a look at the comment section in the biginning portion of the code (line 86 and on)

File attached: HWSCRT.c (81.7 KB)

So in summary , SteadyState solver is an approximation of the diffusion problem for large diffusion constants. There might be some numerical issues when lambda is very small compared to diffusion constants. As far as comparison with Matlab we have not done this. We benchmarked the code agains FEMLAB. 

We would love to add more sophisticated PDE solver quite to CC3D at some point because Steady-State solvers are just an approximations and I wonder if the discrepancy you see between CC#D code and Matlab is because of the slightly different formulation of the problem (i.e. diffusion equation vs Helmholtz equation)

Let me know if this is helpful

attachments(redirected):
.\CC3D_allanswered_linked_files\diffusion_steady_state_ext_potential.zip
.\CC3D_allanswered_linked_files\HWSCRT.c

Comment:

content:
Thank you for your reply! It helped me to understand the problem better!

I see the difference between those two possibilities. I believe my tests rely on the second option: I am solving the equation D(d²u/dx² + d²u/dy²) - k * u = F(x,y) for a long time to determine the steady state of the solution.
(it is the equation in the Reference Manual but with the Diffusion Constant in the Laplacian, as it did not converge otherwise).

Q41: Can you set axis limits when adding plot windows from steppables.py?

content:
the python manual says that you can change some attributes of the plotting windows in the Player, such as log vs. linear axes; however, there doesn't seem to be a way to set axis limits?  I tried _ylim(0,1000), but, no luck, and I thought I should stop fishing around and just ask.

Thanks!!!

Answer:

content:
Hi Kim, as of now this would be quite hard to accomplish from steppable level but we could add this option to the plotting API. We always assumed that plotting in CC3D provides just bare-bone tools but having the ability to specify axes limits should be added. I added a ticket on our github page

Q42: Week 1-Journal/Code Club- PKD Paper

content:
Hey I wanted to do a weekly journal/code club on the PKD paper since Yuan also had questions on it, and it's the new year!


Week 1- 

• Paper: Virtual-Tissue Computer Simulations Define the Roles of Cell Adhesion and Proliferation in the Onset of Kidney Cystic Disease
• Code

• The code from the paper is split into 3 folders: IsoCyst, Measure, and tubule. 
• Tubule- creates a kidney tubule with domains and sub cellular parts
  • PKD.py- This imports models and is where you change and input parameters that feed to the Steppables
  • Lines 1-8: Parameters of interest.

    • #PARAMETERS:
      cd=12          #Typical cell diameter
      Lx=10*cd       #Size of lattice - x
      Ly=10*cd       #Size of lattice - y
      Lz=15*cd       #Size of lattice - z
      Trelax=400     #Time for cells to relax boundaries
      debugFreq=100

      • These parameters are the cell diameter and lattice spacing. I believe they correspond to table S3, since this code is making preformed tubules.  
    • Line 10-13 

      • #POTTS PARAMETERS:
        Temp=50               #Potts temperature
        Time=1+20000             #Total number of MCS
        NOrder=4              #NeighborOrder

        • Temp or Tm is in the Potts formula Eq 2
          • 

             

          • T is from the Ising model, which is physics based. More Tm means more interactions I believe, "To mimic cell motility the method iteratively attempts to move the interfaces between adjacent cells, depending on the amplitude of active membrane fluctuations (expressed as a “cellular temperature” [ μ(τ)) and on a force balance of the active forces the cells exert on their environment (e.g. due to chemotaxis or random motility) and the reactive adhesive, cohesive and cellular compression forces. " Source
          • Norder is the number of voxel neighbors that contact energy is calculated over
          • Similar to lines 15- 17 

            • #CONTACT PARAMETERS:
              CNOrder=4             #Distance of interaction
              CINOrder=4            #Internal Distance of interaction

    • Lines 19-20 

      • #VOLUYME PARAMETERS
        LamV=4                #Lambda Volume
        targV=cd*cd*cd        #Target Volume

      • 
      • so target is the Vt parameter here in the volume constraint part of the GGH equation. 
    • Lines 31-37

      • #CONTACT INHIBITION:
        gFactor=0.8          #Growth factor
        hillCoef=20          #Hill coeficient
        critAlpha=0.35        #Critical alpha
        gFactorSick  =gFactor
        hillCoefSick =hillCoef
        critAlphaSick=critAlpha

      • See the cell contact fraction, under contact inhibition, in the paper for the parameters, and table S4.
        • here K= gfactor which is the max growth rate of the cells (only 1 cell type in this model with 3 sub cellular compartments),ac=critAlpha which is the critical fractional area of cell-cell contact for growth inhibition , n=hillCoefSick. I don't know why critAlphSick and critAlpha are defined separately. The "Sick" is a cell type?
    • Lines 39-41

    • #LUMEN GROWTH:
      Kl=1               
      ApArea=6.*cd*cd*.2  # ApArea=1/Kr

      • Lumen growth, which the tubule module creates in the other stoppable code. Table s2 has info. The growth rate is in volume/time, so voxel/mcs or um3/min. This could be a way to relate experimental results to simulation results.  I think they use this paper from Engelberg as a start for the lumen growth rates (I am trying to figure out which reference Engelberg uses in turn for lumen growth rates). 
  • PKDSteppables.py

attachments(redirected):
.\CC3D_allanswered_linked_files\Screen Shot 2018-01-04 at 2.48.31 PM.png
.\CC3D_allanswered_linked_files\Screen Shot 2018-01-04 at 2.59.37 PM.png

Comment:

content:
Good job！ I have benefited a lot especially for the source of the reference.

Q43: Editing the chemotaxis plugin

content:
I recently read an article which described how cell were first attracted to a chemoattractant, but were later repulsed by the same chemoattractant if the gradient of the field was too steep.
It seemed to me that this could be easily modeled by adding an if-statement in the chemotaxis plugin, which would make the lambda in the chemotaxis module negative if the gradient was steep enough.
But the xml language doesn't use if-statements and I have trouble finding out how to edit the python code of the chemotaxis plugin.
Does anyone know how to edit the chemotaxis plugin or if there is another way of first attracting and then repulsing cells?

Answer:

content:
I don't think editing the plugin is the way to go.

Instead, I would make two cell types, one that chemotaxes up the gradient and another that chemotaxis down the gradient. When the gradient exceeds a limit you switch from one cell type to another. Attached is a quick CC3D project with a fixed cell that is the source of a gradient ("hormone") and a moving cell that switches between going up and down the gradient based on the steepness of the gradient. To calculate the gradient I used the difference between the minimum and maximum gradient values along the moving cell's perimeter. (There might be a more efficient way to do this.) If the gradient is above a threshold the cell switches to going down the gradient, when the gradient is below a threshold the cell switches to going up the gradient. (I takes a while before the cell starts to move since the field has to build up.)

You could also switch the direction based on the field value at (for example) the center of the cell, instead of the field gradient across the cell as I did here.

File attached: InvertChemoract.zip (10.01 KB)

File attached: InvertChemotract.mov (296.64 KB)
Left: Colored by cell type, green is moving up the gradient and blue is moving down the gradient. Right: The "hormone" field created by the fixed cell in the upper right corner. Click the attached .mov to see the animation.

attachments(redirected):
.\CC3D_allanswered_linked_files\InvertChemoract.zip
.\CC3D_allanswered_linked_files\output_movie.png
.\CC3D_allanswered_linked_files\InvertChemotract.mov
.\CC3D_allanswered_linked_files\InvertChemotract.mov

Comment:

content:
Thank you very much for your answer!
I will look into it.

Q44: Smooth Surface of a Cell

content:

Hi,
I'm trying do create a 3D morphogenesis Simulation of the urothelium. I'm draw more or less spehrish cells like this:

        for xr in range(xStart, xEnd):
            for yr in range(yStart, yEnd):
                for zr in range(zStart, zEnd):
                    rd = sqrt((xr - x0) ** 2 + (yr - y0) ** 2 + (zr - z0) ** 2)
                    if (rd <= radiusPx):
                        steppable.cellField[xr, yr, zr] = cell

With this function I'm able to get some sperish cells, but without a smooth surface (see first picture).
Furthermore, during the growth of a cell, they have sometimes pixels outside of the cell (see second picture). I place the cells just somewhere in the lattice, in the second picture, to see more details of the growth process.
Since, I want to create sperish cells I use formulas of a sphere to calculate the TargetVolume and the TargetSurface (the TargetSurface is doubled, so that the cell will growth). I use the following lambda values: LambdaVolume = 1.0 and LambdaSurface = 2.0

Do you know any tips or tricks about how to get a smooth surface on a cell?

Thank you

attachments(redirected):
.\CC3D_allanswered_linked_files\points outside of a cell.png
.\CC3D_allanswered_linked_files\Screenshot at 2018-01-08 07-56-38.png

Answer:

content:
A couple suggestions;
* In the Potts section, set the NeighborOrder to 5 to reduce pinning to the lattice.
* The rough surface of the cells is probably caused by the targetSurface being 2X bigger than it should be for the current cell target volume. Try increasing both the target surface and target volume properly for the increasing radius. (Or, if you are changing the volume, recalculate the surface every time.)  Instead of setting the targetSurface much too big simply reduce the lambdaSurface significantly.

if cell.targetVolume <= cell.volume:
    cell.targetVolume += 1
    radius = (3./4./3.14159*cell.targetVolume)**(1./3.)
    cell.targetSurface = 4.*3.14159*radius**2 * 1.1  # 10% extra surface​

Comment:

content:
Thanks,
the drawn cells look way better.

Since it is not possible to provide another picture in the comment I use the answer function to show them!

Answer:

content:
One problem I still have is that during the simulation become cubic again. I just checked the volume - TargetVolume and the surface - TargetSurface values and saw that the Surface is way to large regarding the volume value (see second picture, at the console the line at the cursor position).

Every MCS the TargetVolume and the TargetSurface, regarding TargetVolume, are calculated, with the provided formula.
I tried to simulate with LambdaVolume=1.0 and LambdaSurface=0.2 and 0.05 (the pictures are of the second simulation with LambdaSurface=0.05).

Since there is no way to set the surface value of a cell, are there other possibilities to influence these values of a cell than to calculate the TargetSurface and set the LambdaSurface? Where could those large surface values come from?

attachments(redirected):
.\CC3D_allanswered_linked_files\Screenshot at 2018-01-09 11-02-48.png
.\CC3D_allanswered_linked_files\Screenshot at 2018-01-09 11-03-49.png

Answer:

content:
It might be due to your adhesion energies. For example, if the cell's surface is at the target surface value and then the cell grows by a pixel (or voxel) then you often add 2 (in 2D, 4 in 3D) to the surface and that number of values of the cell to medium adhesion. So adhesion energy will play into the surface and volume behavior of a cell.

Another thing that makes getting spherical cells tricky is that the surface on a square lattice is always greater than for a real sphere. You need to have a somewhat larger targetSurface than the simple volume vs. surface equations for a sphere.

One thing you might consider is doing a parameter scan to find the set of parameters that works best for your size cells. I would scan targetVolume, lambdaVolume, LamdaSurface, the adhesion j values and a parameter/function that converts targetVolume into a suitable targetSurface, perhaps also the Potts temperature.

Comment:

content:
Thanks, these are valuable facts.
Your right with the second point. I have to find a value where the TargetSurface of cell is not too larger, as it was with the initial problem, and not too small.

I tried different adhesion values but in a range of 0 to 50, for every adhesionValue in the energy Matrix it did not change the behavior of the cells, as they still become cubes sooner or later.

I wonder how the initial Volume and Surface is calculated:
The cells are drawn with a radius of 4px:
    Therefore the Volume of the cube around the cell is 512vx³ and the surface are 384vx²
    Whereas the Volume of the drawn sphere should be 268vx³ and the surface 201vx²
CC3D calculates a Volume of around 255(vx³ - should be the right unit) and a surface of around 283vx² -> The given values vary a bit with +-2
   Why can the given surface value already be larger than the volume? Moreover the surface value is not near the surface value of the cubic neither the surface, how does this happen?

Q45: How to specify the values of the parameters in GGH model?

content:
Dear all, there are some puzzles for GGH model.
1. How to specify the values of parameters in the model? For example, adhesion energy J, elastic constants K. Can I set them according to the experiments? How to realize?
2. As for the time, the Monto Carol Step(MCS) is determined by the real-time. However, the values of the parameters in simulation based on the relationship of MCS and real-time and the values have an impact on the total MCSs, i.e. they are coupled. How to solve this?
3. How to specify the relative values of the parameters?
Thank you~

Answer:

content:
Hi Yuan,

If I understand the question correctly, the way I map MCS to real time is to look at the average square displacement of cells in experiments and simulations and then use this to directly map MCS to real time.

Mapping adhesion energy to experiments is more complicated, unless you can extract real values or find values in literature, it is very difficult, one possible way is to use ratios depending on the minimization or maximization of cell boundary interfaces, i.e if two cells in a non confluent environment in contact increase their shared contact area they more adhesive than cells that decrease their contact area.

best,

Daniel

Comment:

content:
Hi Daniel,

Thanks for your answer. I still have two questions:

1. The MCS and real-time relationship can only be obtained after the simulation and experiment?

2. How to specify the relative size of the parameters?

Thank you~   

Yuan

Answer:

content:
The mcs to time relationship can often be determined simply by inspecting your problem. How fast do things happen? Are cells moving and/or dividing? How long does it take to move or divide? You mcs to time relationship is typically set so that an mcs represents a small portion of your time domain. If it takes 2 hours for a cell to divide then a 1 second/mcs time scale might be OK since that means there are 7200 mcs over the course of the process, which is likely to be sufficient.

Alternatively, you can scan the mcs to time relationship. Compare the results of simulations with 1 mcs= 1 hour, 1 minute, 1 second, 1 millisecond, ... Usually there will be a range of time scales where the simulation output changes with the time scale and another range of time scales where the output doesn't change. You want your simulation to be running in the time step size independent domain.

Q46: Week 2- Journal/Code Club- PKD Paper

content:
Week 2-
Paper: Virtual-Tissue Computer Simulations Define the Roles of Cell Adhesion and Proliferation in the Onset of Kidney Cystic Disease
Code-  http://www.compucell3d.org/Models 

Discussion on Code/Basics-Focusing on Tubule code
The code from the paper is split into 3 folders: IsoCyst, Measure, and tubule.

• Tubule- creates a kidney tubule with domains and sub cellular parts
  • PKD.py- This imports models and is where you change and input parameters that feed to the Steppables
    • Continuing from Line 42

    • def configureSimulation(sim,Lx,Ly,Lz,Temp,Time,NOrder,CNOrder,CINOrder,debugFreq):

      • This defines the class configureSimulation. This class is used to import XML type info and parameters for the model. It is called called later in the same script on line 169, see below:

        • configureSimulation(sim,Lx,Ly,Lz,Temp,Time,NOrder,CNOrder,CINOrder,debugFreq))

    • Lines 44-45

      •  import CompuCellSetup
            from XMLUtils import ElementCC3D

      • This imports the utilities so the XML import can work
    • Line 47

      • CompuCell3DElement=ElementCC3D("CompuCell3D",{"Version":"3.7.3"})

      • This creates a CC3D element or a CC3D kernel, which is the way you do XML import. Specifically, it creates a top level element, or a root element, using the function Elementcc3d() from XMLUtils. See chapter 17 of the 3.7.6 version of Python Scripting manual 
    • Lines 49-55

      • #POTTS CONDITIONS
            potts=CompuCell3DElement.ElementCC3D("Potts")
            potts.ElementCC3D("Dimensions",{"x":Lx,"y":Ly,"z":Lz})
            potts.ElementCC3D("Temperature",{},Temp)
            potts.ElementCC3D("Steps",{},Time)
            potts.ElementCC3D("NeighborOrder",{},NOrder)
            potts.ElementCC3D("DebugOutputFrequency",{},debugFreq)
          #Setting periodic boundary conditions to all directions
            potts.ElementCC3D("Boundary_x",{},"Periodic")
            potts.ElementCC3D("Boundary_y",{},"Periodic")
            potts.ElementCC3D("Boundary_z",{},"Periodic")

      • These lines import XML to the CC3D kernel created before. Specifically, Line 47 created a high level element, and now we are attaching child elements to it (here it is the Potts child element).  Potts itself has child elements, so we attach child elements to Potts element
    • Lines 62-70

      • #CELL TYPES -> Lumen, Cyto, Apical, Basal, Lateral1, Lumen and Lateral2
            cellType=CompuCell3DElement.ElementCC3D("Plugin",{"Name":"CellType"})
            cellType.ElementCC3D("CellType",{"TypeName":"Medium","TypeId":0})
            cellType.ElementCC3D("CellType",{"TypeName":"Cyto","TypeId":1})
            cellType.ElementCC3D("CellType",{"TypeName":"Apical","TypeId":2})
            cellType.ElementCC3D("CellType",{"TypeName":"Basal","TypeId":3})
            cellType.ElementCC3D("CellType",{"TypeName":"Lateral","TypeId":4})
            cellType.ElementCC3D("CellType",{"TypeName":"Lumen","TypeId":5})
            cellType.ElementCC3D("CellType",{"TypeName":"Lateral2","TypeId":6})

      • We are importing more XML. Here it is the Celltype plugin. I am trying to figure out what the two Lateral cell types are for.

Q47: Cell compression by plates.

content:
Dear all,
I am trying to reproduce the process of cell compression, as shown below(picture and PP110 in PDF). There is only one big cell for simplicity. An energy term  $E=\frac{1}{2}k\left(z-z_0\right)^2$E=12 k(z−z₀)²  is introduced to represent the effect of the plates.  $z$z  is the vertical position,  $z_0$z₀  the 0 energy position of the plate interface and  $k$k  the stiffness of plate.
I manipulate the energy term in python referring to the manual,
 https://cc3dadvancedtuts.wordpress.com/2015/08/01/manipulating-the-energy-term-for-a-pixel-flip-attempt-in-python/
I am meeting some difficulties in realizing the algorithm.
First, the vertical position  $z$z is obtained by calculating the average value of the outmost layer pixels of the plates. I try to use the boundary tracker plugin, but I am not sure whether it can be used in the python plugin.
What's more, the changed energy is  $\bigtriangleup E=\frac{1}{2}k\left(z_{n+1}-z_0\right)^2-\frac{1}{2}k\left(z_n-z_0\right)^2=k\left(\frac{z_{n+1}+z_n}{2}-z_0\right)\left(z_{n+1}-z_n\right)$△E=12 k(z_n+1−z₀)²−12 k(z_n−z₀)²=k(z_n+1+z_n2 −z₀)(z_n+1−z_n) . It seems that the changed energy is related to the vertical position before and after the pixel flip. How to realize it ?
I have a try and the outcome seems not positive, the imperfect model is attached.
Thank you!

Yuan
File attached: version_finale.pdf (6.24 MB)

attachments(redirected):
.\CC3D_allanswered_linked_files\QQ截图20180115161521.png
.\CC3D_allanswered_linked_files\compressionCell.rar
.\CC3D_allanswered_linked_files\version_finale.pdf

Answer:

content:
In the figure (labeled 5.7) it doesn't appear that the plates are deformed, only the cell deforms. Do you really want the plates to deform or just the cell?

Looking at your CC3D, there are a couple things if you want the plates to be deformable;
1. You need to define the plates as cells with target volumes and probably target surfaces.
2. When you move the plates (or enlarge the plates) you need to run enough MCS to let the system relax between movements.

If you really want deformable plates you might be able to implement a finite element mesh in CC3D for the plates. The plates would be a set of small cells coupled by springs to their neighbors using focal point plasticity links.

Comment:

content:
Actually，in the figure the deformation is small, but I think it's okay to neglect the deformation of plates since they are much stiffer than cells.
My question is, are there other methods to move the plates as rigid bodies?

Comment:

content:
You can define the plates as fixed cells then add a row of pixels to those fixed cells in python, then relax the system for some number of MCS, then add another row of pixels, ...

In twedit++ use the shortcut at;
   CC3D Python -- Visit -- All Lattice Pixels
for how to access all the lattices pixels

You can assign a particular pixel to a cell ("cellA") with;
  pt.x=4; pt.y=6; pt.z=0
  self.cellField.set(pt,cellA)
or
  self.cellField.set((x,y,z),cellA)

The pixels might have been medium or your cell being squashed. That is why you'll need to run some number of MCS to relax things once you've enlarged the wall.

Q48: Periodic boundary condition for a hexagon

content:
Dear all, 
In a CPM simulation of Drosophila retina cells, the authors apply a periodic boundary condition for hexagons. As shown below, the cells marked with the same symbol are treated as part of the same cell. Can I realize this using CC3D? Thank you!
Best,
Yuan

attachments(redirected):
.\CC3D_allanswered_linked_files\QQ截图20180116163449.png

Answer:

content:
I believe CC3D will only do periodic boundaries on a rectangular bounding box.

Comment:

content:
So, is it convenient to make changes in the source code or do I have to program by myself?

Q49: Some puzzles on concepts.

content:
Dear all,
I meet some puzzles on some concepts in different papers.
1. the "source"(expanding) cell and "destination" (being overwritten) cell in CPM described in the reference manual and the pixel-copy-source and pixel-copy-destination pixel.
Are the positions of source and destination opposite for cells and pixels? That's to say, the pixel-copy-source pixel belongs to the "destination"(old) cell and the pixel-copy-destination pixel belongs to the "source"(new) cell? For pixels, do the "source" and "destination" in the manual have the same meaning of the "source" and "target" in the following color picture from other paper?  

Thank you~
Yuan

attachments(redirected):
.\CC3D_allanswered_linked_files\图片1.png
.\CC3D_allanswered_linked_files\QQ截图20180117105653.png
.\CC3D_allanswered_linked_files\QQ截图20180117105623.png
.\CC3D_allanswered_linked_files\QQ截图20180117105623.png
.\CC3D_allanswered_linked_files\QQ截图20180117111049.png

Answer:

content:
Actually this is the correct way of associating source destination old and new:

the pixel-copy-source pixel belongs to the "source"(new) cell and the pixel-copy-destination pixel belongs to the "destination"(old) cell?

Q: For pixels, do the "source" and "destination" in the manual have the same meaning of the "source" and "target" in the following color picture from other paper?  

A: It appears that the answer is no. To me source is the pixel of that belongs to a cell (new cell) that will try to overwrite destination (old cell) . This is how things are coded in CC3D

Does this make sense?

Comment:

content:
Yeah, I see. There are some differences though both of them are CPM models. Besides, is there a manual explaining the implementation of the CC3D code？  

Answer:

content:
Yes, it is for older version 3.4.1, however. Still the core concepts are there and architecture of plugins and steppables has not changed much - in particular w.r.t what is new(source ) cell and what is old(destination) cell

http://www.compucell3d.org/BinDoc/cc3d_binaries/Manuals/Developers_Documentation_v3.4.1.pdf

Take a look at it

Q50: Parameter scan for parameters in a python file

content:
I would like to run parameter scan from parameters listed in the .py file of the Main Python Script. However, I am unable to open the editor when I click on the parameters to be changed. I guess then I will have to input it manually in the xml file. I would highly appreciate if I can avail a template for the same.

Answer:

content:
Can you move the parameters into your steppables python file instead of your main python file? The parameter scan is, I believe, specific to regular python files and to the CC3D XML file.

See also: http://pythonscriptingmanual.readthedocs.io/en/latest/parameter_scans.html

Comment:

content:
Do I have to make them global as well  as mentioned here?

from PySteppables import *
import CompuCell
import sys

MYVAR = 10
MYVAR1 = 'new str'


class CellSortingSteppable(SteppableBasePy):
def __init__(self, _simulator, _frequency=1):
SteppableBasePy.__init__(self, _simulator, _frequency)

def step(self, mcs):
global MYVAR
print 'MYVAR=', MYVAR
for cell in self.cellList:
if cell.type == self.DARK:
cell.lambdaVecX = -0.5

Answer:

content:
I believe they need to be in the top level namespace (hence "global"). In the little demo snippet the "global MYVAR" line isn't needed if the CellSortingSteppable class is only reading, and not writing, the variable.  All of your classes will have direct access to any variable defined outside the classes.

You can first try it without making the variables global (e.g., a variable in a class), it might work.

Q51: Order of parameters in parameter scan

content:
I am confused about what is the order in the which the parameters are chosen from the list of parameter scan. Are all parameters varied at the same time randomly or is there a sequence for the variation. I have 8 parameters to be varied and I chose 5 values for each. If I am thinking right, I should get 0 , 1 , 2, 3 ,4 and 5 directories in the output folder. However, how would I be able to know the results correspond to which values of parameters ?

Answer:

content:
The parameter scan task exhaustively enumerates all possible combination of your parameters. 8 parameters with 5 values each is equivalent to an 8-digit base 5 number so there are 5^8=390,625 simulations. If it takes 10 minutes per simulation, then it will take 2,712 days to do them all. As you can see, parameter scanning can rapidly spin up into an intractable number of simulations.

The order of simulation is the same as outlined in the scan xml file. If you have 3 parameters (A,B,C) and are doing 2 values (1,2) each, the sequence (top to bottom in the xml file) will be;

A1B1C1
A2B1C1
A1B2C1
A2B2C1
A1B1C2
A2B1C2
A1B2C2
A2B2C2
And there are 2^3=8 total simulations.

CC3D doesn’t do anything in particular to help you tract the results. All the output files are created, and each set is in its own numbered directory. You can examine each of the directories and see what values were used and what the results are but that isn’t very easy.

 A better approach is to add code to your model to log all the simulation results into a single output file. You would log the parameters used for a particular run along with one or more results. Can you define a few numeric metrics for a simulation? Perhaps the number of cells at the end of the run, or the number of cells that died during the run. Or, if you are fitting experimental data, a residual error or chi-squared type metric describing how well the particular run reproduced the experimental data. To actually log that data, and preferable to have all the simulation log into the same file, you can add code to a “finish” block of one of your steppables to write the file at the end of the run. Below is sample code to give you an idea how to do it.

def finish(self):  # part of one of your python steppables

    # Finish Function gets called after the last MCS

    #   

    # Write a small summary of parameter scanning to a file.

    # Since we want all the simulaltions writing to the same file we

    # need it in an obvious and unambigous place. Here we assume we can just write

    # to the Windows’ root directory of c:\

 

    # Buld up a tab-delimited stirng with the scan paramters actually used in this run plus

    # some measures of the simulation’s results

    # myParam1,myParam2, ... are the variables being scanned and are probably global

    # myOutput are the result of some calculation, the number of cells, whatever you want to log

    outLine  = str(myParam1)+"\t"+str(myParam2)+"\t"

    outLine += str(myParam3)+"\t"+str(myParam4)+"\t"

    outLine += str(myOutput1)+"\t"+str(myOutput1) # add as many items as you need

    outline += "\n" # need a trailing line break!

   

    #open a file for append, write the string and close the file

    logfile = open(“C:\\ParmaterScan_log”,'a+') # the ‘a+’ means create if it doesn’t exist,

    logfile.write(outLine)                      # append to it otherwise

    logfile.close()​

Q52: Linear Gradient

content:
I was wondering if there is any way to create a linear gradient for chemotaxis.
So far I have tried using a second cell type to secrete a chemoattractant and I managed to make the edges of the window secrete by giving the boundary conditions in the diffusion solver steppable code a value.
But this produces a gradient which is steeper at the sides than in the middle
.

Does anyone know a way to make the gradient linear?

attachments(redirected):
.\CC3D_allanswered_linked_files\Gradient.png

Comment:

content:
Hi Tarqwin!  You should be able to do this by making the boundary conditions not constant, but a zero derivative (zero change at each point along the boundary, so each pixel-length of boundary will stay at the same value as the pixel next to it). The code below is the chunk for a complete field inside of your diffusion solver of choice (I am using DiffusionSolverFE).  The constant boundaries portion is commented out.  I assume you will not need the z-dimension boundaries, either....

Hope this works for you!

  <DiffusionField Name='Chemical'>
    <DoNotScaleSecretion/>
    <DiffusionData>
      <GlobalDiffusionConstant>1.0</GlobalDiffusionConstant>
      <GlobalDecayConstant>0</GlobalDecayConstant>
    </DiffusionData>
    <SecretionData>
      <ConstantConcentration Type="Cell">0.0</ConstantConcentration>
    </SecretionData>
    <BoundaryConditions>
       <!-- CONSTANT BOUNDARIES -->
      <!-- <Plane Axis="X"> -->
      <!--   <ConstantValue PlanePosition="Min" Value="0"/> -->
      <!--   <ConstantValue PlanePosition="Max" Value="0"/> -->
      <!-- </Plane> -->
      <!-- <Plane Axis="Y"> -->
      <!--   <ConstantValue PlanePosition="Min" Value="0"/> -->
      <!--   <ConstantValue PlanePosition="Max" Value="0"/> -->
      <!-- </Plane> -->
      <!-- <Plane Axis="Z"> -->
      <!--   <ConstantValue PlanePosition="Min" Value="0"/> -->
      <!--   <ConstantValue PlanePosition="Max" Value="0"/> -->
      <!-- </Plane> -->
      <!-- <Plane Axis="X"> ></Plane> <Plane Axis="Y"><Periodic/> </Plane> -->
       <!-- CONSTANT DERIVATIVE BOUNDARIES -->
        <Plane Axis="X">
          <ConstantDerivative PlanePosition="Min" Value="0"/>
          <ConstantDerivative PlanePosition="Max" Value="0"/>
        </Plane>
        <Plane Axis="Y">
          <ConstantDerivative PlanePosition="Min" Value="0"/>
          <ConstantDerivative PlanePosition="Max" Value="0"/>
        </Plane>
        <Plane Axis="Z">
          <ConstantDerivative PlanePosition="Min" Value="0"/>
          <ConstantDerivative PlanePosition="Max" Value="0"/>
        </Plane>
        
      </BoundaryConditions>
    </DiffusionField>​

Answer:

content:
Do you want a static chemical gradient or a changing one? If you want a static gradient, you can set the initial condition to be, e.g. x* constant, and then set the diffusion constant and decay constants to zero. 
e.g.

<Steppable Type="DiffusionSolverFE">
      <DiffusionField Name="Oxygen">
         <DiffusionData>
            <FieldName>Oxygen</FieldName>
            <DiffusionConstant>0.0</DiffusionConstant>
            <DecayConstant>0.0</DecayConstant>
            <InitialConcentrationExpression>x*CONSTANT</InitialConcentrationExpression> 
         </DiffusionData>
      </DiffusionField>
</Steppable>​

If you don't want the gradient to be static, because you will do something else to it later, you can use fixed boundary conditions along the X=0 and x=x_max edges and periodic boundaries along the other sides.

<Steppable Type="DiffusionSolverFE">
    <DiffusionField Name="Oxygen">
         <DiffusionData>
            <FieldName>Oxygen</FieldName>
            <DiffusionConstant>5.0</DiffusionConstant>
            <DecayConstant>0.001</DecayConstant>
            <InitialConcentrationExpression>x*10/x_dim</InitialConcentrationExpression>
         </DiffusionData>
         <BoundaryConditions>
            <Plane Axis="Y">
               <ConstantValue PlanePosition="Min" Value="0.0"/>
               <ConstantValue PlanePosition="Max" Value="10.0"/>
            </Plane>
            <Plane Axis="X">
               <Periodic/>
            </Plane>
         </BoundaryConditions>
      </DiffusionField>
   </Steppable>

I don't have CC3D running on this computer, so check that I haven;'t reversed X and Y!

Comment:

content:
This is the way to do it. You may also set concentration directly in Python but the solution that James presented should do the trick

Comment:

content:
Can you specify a gradient in specific regions on a lattice? For example, paper-discs with antibiotic on a petri dish? I imagine it would involve specifying boundary conditions in specific regions.

Answer:

content:
I would add it as a new cell type and then add a diffusing chemical to that cell type. You would initialize your "disc" cell, perhaps with BlobInitializer. Then, assuming reduction in the available antibiotic diffusing from the disc is negligible compared to the length of the simulation, you would set a constant concentration of "antibiotic" for your "disc" cell type in your diffusion solver above. No need to change boundary conditions from what James said above (sorry I missed putting in the constant boundary for your previous question).

Comment:

content:
Ah ok, that makes sense. I will try that. I have cell layers, so that will work if I put the source underneath at z=0 and put the cell layers at z>1, for a 2D model at least.

Comment:

content:
yes, that might work if you embed the source into the z=0 boundary, using the boundary definitions. I'm not sure if you can put cells (in this case the disc) into the boundary level; I also have never tried defining a small area of a boundary as constant -- you might also have to define the rest of the plane as 0, or something like that. If you don't want to play with defining constant conditions for all of the z boundary, and instead add the disc cell type, I think it will take up that space in your 2-D domain, and cells will not crawl on top.... however, if your disc is small, I doubt that it would change the result of your simulation (and it would be easier to implement, most likely!).

Q53: Counting Cell Cycles

content:
System: macOS Sierra v 10.12.6
CC3D version: 3.7.6

I have a simulation where I have two cell types: dividing and non dividing. 

<Plugin Name="CellType">
      <CellType TypeId="0" TypeName="Medium"/>
      <CellType TypeId="1" TypeName="dividing"/>
      <CellType TypeId="2" TypeName="nondividing"/>
</Plugin>​

I am also using the Growth and Mitosis steppables. The simulation starts off with a dividing cell that will grow to a certain size and then have a certain probability of the daughter cells remaining as dividing cells or become non dividing cells. This will repeat for how ever many MCS I chose.  

My goal is to try to determine how many times a cell divides before it becomes a nondividing cell. I thought about using the cell.dict like the manual suggested but I haven't had success.  How can I get a dictionary (or a list or something similar) containing cell ids and the number of times it divided?  

from PySteppables import *
import CompuCell
import sys
import random
import numpy as np
from PySteppablesExamples import MitosisSteppableBase

class ConstraintInitializerSteppable(SteppableBasePy):
    def __init__(self,_simulator,_frequency=1):
        SteppableBasePy.__init__(self,_simulator,_frequency)
    def start(self):
        cellcell = self.newCell(self.DIVIDING)
        self.cellField[100:105, 100:105, 0] = cellcell
        cellcell.targetVolume=36
        cellcell.lambdaVolume=2.0 

class GrowthSteppable(SteppableBasePy):
    def __init__(self,_simulator,_frequency=1):
        SteppableBasePy.__init__(self,_simulator,_frequency)
    def step(self,mcs):
        for cell in self.cellListByType(self.DIVIDING):
            cell.targetVolume+=0.1 

class MitosisSteppable(MitosisSteppableBase):
    
    def __init__(self,_simulator,_frequency=1):
        MitosisSteppableBase.__init__(self,_simulator, _frequency)
        
    def step(self,mcs):
        cells_to_divide=[]
        for cell in self.cellList:
            if cell.volume > 72:
                cells_to_divide.append(cell)
        
        for cell in cells_to_divide:
            self.divideCellRandomOrientation(cell)

    def updateAttributes(self):
        roll_dice = random.random()
        self.parentCell.targetVolume /= 2                 
        self.cloneParent2Child()
        
        for cell in self.cellList:
            if roll_dice > 0.3 :             
                self.childCell.type = self.DIVIDING
                self.parentCell.type = self.DIVIDING
            else:
               self.childCell.type = self.NONDIVIDING
               self.parentCell.type = self.NONDIVIDING
        

Answer:

content:
As you suspected the dictionary is one way to do it. When you create a cell at the start of the simulation and every time the cell divides you can increment a counter in a dictionary associated with the cell. When you create the cell (the last line is the dictionary entry);

    def start(self):
        cellcell = self.newCell(self.DIVIDING)
        self.cellField[100:105, 100:105, 0] = cellcell
        cellcell.targetVolume=36
        cellcell.lambdaVolume=2.0
        cellcell.targetSurface=24
        cellcell.lambdaSurface=2.0
        cellcell.dict['age'] = 1  ### The "age" of the cell, increment every time the cell divides​

    def updateAttributes(self):
        roll_dice = random.random()
        self.parentCell.targetVolume /= 2                 
        self.cloneParent2Child()
       
        for cell in self.cellList:
            if roll_dice > 0.3 :            
                self.childCell.type = self.DIVIDING
                self.parentCell.type = self.DIVIDING

            else:
               self.childCell.type = self.NONDIVIDING
               self.parentCell.type = self.NONDIVIDING

        self.childCell.dict['age'] += 1   ### The "age" of the cell, increment everytime the cell divides
        self.parentCell.dict['age'] += 1  ### The "age" of the cell, increment everytime the cell divides​

        if self.mcs % 100 == 0:         ###  only report every 100 mcs
            type1=0;   type2=0          ###
            print "\nmcs=",mcs          ###
            for cell in self.cellList:  ###
                if cell.type == 1:      ###
                    type1 +=1           ###
                if cell.type == 2:      ###
                    type2 +=1           ###
                print "cell id,type,age",cell.id,cell.type,cell.dict['age']  ###
            print "Type1=",type1,"  Type2=",type2,"\n" ###​

attachments(redirected):
.\CC3D_allanswered_linked_files\divide_nodivide_track.png
.\CC3D_allanswered_linked_files\divide_nondivide_track.zip

Comment:

content:
Thank you! This did the trick. 

Q54: 3D Cell Shapes

content:
Hi,

I wonder if it is possible not only to draw 3D shapes but also keep them during growth and mitosis events?
Does lay around every cell a rectangular box or is it possible to select a different shape for a 3D simulation?

Thank you,
Thorsten

Answer:

content:
your best bet if you want to have control over cell shape is to use compartmentalized cells.

See 

http://www.compucell3d.org/BinDoc/cc3d_binaries/Manuals/CompuCell3D_Reference_Manual_v.3.7.4.pdf

page 27

and check chapter of length constraint plugin that follows. 

 As far as mitosis this should help you get started:

http://pythonscriptingmanual.readthedocs.io/en/latest/dividing_clusters.html

Comment:

content:
To add more to my previous answer. The shape of a single cell could be controlled by choosing properly contact energy (NeighborOrder parameter in the Contact plugins plays actually and important role so make sure it is not 1 but bigger to avoid rectangular lattice-related artifacts). Additionally you would control contact energy coefficients and also you may include surface and volume constraints. You may also play with Length constraints. I would suggest that when doing 3D simulation, to start with a single cell and get the parameters right first. Then slowly keep adding components to your simulation and see how things behave. For example you could study what happens at the mitosis stage starting with just one cell. Also to mak simulation run faster start with smaller lattice and increase the size as you add more components. this way you will not waste time by probing the lattice regions that have no cells. Let me know if this answers your questions

Comment:

content:
Thanks for the extensive answer. It definitely helped.

I'm still wondering how the volume and surface of each single cell is calculated?
Is it the calculation for a cuboid, since in 3D we have voxels to display, or is more deeply and than the amount of lattices sites within each cell?

Comment:

content:
In a 2D CC3D model the "cell surface" is actually a length and the "cell volume" is actually an area. The units are pixels and pixels^2, respectively.

In a 3D CC3D model the "cell surface" is a true area and the "cell volume" is a true volume. The units are pixels^2 and pixels^3, respectively.

Area and volume, in both 2D and 3D, are always integers.

In both 2D and 3D the measurement is over the actual set of pixels (or voxels) and no smoothing or interpolation of the surface is done. As a result, the surfaces tend to be larger than they would be on a smooth surface since the lattice based surface "staircases".

Q55: How to show the time(MCSs) on the screen?

content:
Dear all,
I want to show the simulation time (MCSs) on the screen. How to realize this?
Thank you~
Yuan

Answer:

content:
The MC Step is already shown in the bottom left corner of the CC3D window. Do you want the MC step displayed in the sub-window that has the cell display? As far as I know there isn't a way to do that. If you want to create an animation of a CC3D simulation (from the saved screen shots from CC3D) you can use ImageMagick to add the MC step to the images as they are added to the animation.

Q56: How to simulate the inner force of a cell?

content:
Hi, everyone
I wonder if I can simulate the inner force of a cell, for example, the movement of the spindle body elongates the cell in a certain direction, i.e. the cell grows along a certain direction.
Thank you.
Yuan

Answer:

content:
I would say that anything that requires more control of the subcellular structures, mechanisms etc should be handled via compartmentalized cells. 

Having said this, you could explore LengthConstraint plugin that does precisely what you want - elongates cells. One thing to keep in mind is to make sure you include ConnectivityGlobal plugin. This plugin has been updated in 3.7.7 to run much faster than previous version (especially in 3D) so make sure you are using the latest CC3D if you want to go down that route. 
I would check elongation demos in the Demo folder of CC3D install and also there is a contributed module called OrientedGrowth (the corresponding demo is OrientedGrowthDemo) so check them out

Q57: The simulation of cell vibration

content:
Dear all,
I am trying to simulate the process of cell vibration, i.e. the cell's shape size changes in a certain period. What should I do? Thank you!

Answer:

content:
You can keep changing and sorface and volume constraints from the pythn steppable . say at mcs=0

targetVolume=25
at mcs10
targetVolume=26
at mcs20
targetVolume=27
...
at mcs 200
targetVolume=20

...

Not sure if this is what you are after but this gives you an idea how to simulate what you want

Q58: Week 3- Journal/Code/- PKD Paper, PKD.py lines 70-end

content:
Week 3-  Picking up from last week at line 70 of PKD.py
Paper: Virtual-Tissue Computer Simulations Define the Roles of Cell Adhesion and Proliferation in the Onset of Kidney Cystic Disease
Code-  http://www.compucell3d.org/Models

Discussion on Code/Basics-Focusing on Tubule code
The code from the paper is split into 3 folders: IsoCyst, Measure, and tubule.
Tubule- creates a kidney tubule with domains and sub cellular partsPKD.py- This imports models and is where you change and input parameters that feed to the SteppablesContinuing from Line 42

• PKD.py-
  • lines 72-82

  • #CELL COLORS
        cellColor=CompuCell3DElement.ElementCC3D("Plugin",{"Name":"PlayerSettings"})
        cellColor.ElementCC3D("Cell",{"Type":0,"Color":"#000000"}) # Black        -> Medium
        cellColor.ElementCC3D("Cell",{"Type":1,"Color":"#FFFF99"}) # Light yellow -> Cyto
        cellColor.ElementCC3D("Cell",{"Type":2,"Color":"#009900"}) # Dark Green   -> Apical
        cellColor.ElementCC3D("Cell",{"Type":3,"Color":"#990000"}) # Dark red     -> Basal
        cellColor.ElementCC3D("Cell",{"Type":4,"Color":"#FF0000"}) # Red          -> Lateral1
        cellColor.ElementCC3D("Cell",{"Type":5,"Color":"#00E6E6"}) # Light blue   -> Lumen
        cellColor.ElementCC3D("Cell",{"Type":6,"Color":"#0000FF"}) # Blue         -> Lateral2
        cellColor.ElementCC3D("TypesInvisibleIn3D",{"Types":"0,1,4,5"})   # Don't show Medium,Cyto, Lateral1 and Lumen
        cellColor.ElementCC3D("VisualControl",{"ScreenshotFrequency":250,"ScreenUpdateFrequency":10})
    

    • The first line, line 73, creates a compucell element for the "Playersetting" plugin. The element is a root element object, so we can add the child elements for each color later on. See Chapter 17 pg 56 in the 3.7.6 Python Scripting Manual on the CC3D website for more info.
  • Lines 74-80
    • These lines assign colors to display in the program for each cell type. They are assigned by using the Type number
  • Line 81-82

    •   cellColor.ElementCC3D("TypesInvisibleIn3D",{"Types":"0,1,4,5"})   # Don't show Medium,Cyto, Lateral1 and Lumen
          cellColor.ElementCC3D("VisualControl",{"ScreenshotFrequency":250,"ScreenUpdateFrequency":10})

    • These lines adjust visibility so Medium, Cato, Laterall and Lumen aren't visible on the display. When I run the program I see the Lumen and Medium, so I am not sure if this is overwritten elsewhere.
  • Line 86 

     contact=CompuCell3DElement.ElementCC3D("Plugin",{"Name":"Contact"})​

    • This creates a room element object for the Contact Energy plugin
  • Lines 88-93 

    contact.ElementCC3D("Energy",{"Type1":"Medium","Type2":"Medium"},0)
        contact.ElementCC3D("Energy",{"Type1":"Medium","Type2":"Cyto"},60)
        contact.ElementCC3D("Energy",{"Type1":"Medium","Type2":"Apical"},60)
        contact.ElementCC3D("Energy",{"Type1":"Medium","Type2":"Basal"},3)
        contact.ElementCC3D("Energy",{"Type1":"Medium","Type2":"Lateral"},8)
        contact.ElementCC3D("Energy",{"Type1":"Medium","Type2":"Lateral2"},8)    ​

    • These lines assign contact energies for the cells to the medium. This is the J term in the GGH model equation. There are J's for each cell-cell pair possible, including the medium. The other J's are in code that follows.
  • Lines 95-100 

    #Cyto
        contact.ElementCC3D("Energy",{"Type1":"Cyto","Type2":"Cyto"},60)
        contact.ElementCC3D("Energy",{"Type1":"Cyto","Type2":"Apical"},60)
        contact.ElementCC3D("Energy",{"Type1":"Cyto","Type2":"Basal"},60)
        contact.ElementCC3D("Energy",{"Type1":"Cyto","Type2":"Lateral"},60)
        contact.ElementCC3D("Energy",{"Type1":"Cyto","Type2":"Lateral2"},60)​

    • The J terms for cell-cyto pairs. Note they all are the same. This means the contact energy between cells to the cyto cell compartments is the same, or no preference.
  • Lines 101-105

    #Apical
        contact.ElementCC3D("Energy",{"Type1":"Apical","Type2":"Apical"},15)
        contact.ElementCC3D("Energy",{"Type1":"Apical","Type2":"Basal"},20)
        contact.ElementCC3D("Energy",{"Type1":"Apical","Type2":"Lateral"},20)
        contact.ElementCC3D("Energy",{"Type1":"Apical","Type2":"Lateral2"},20)​

    • For apical-apical, there is a J of 15. compare this to 20 for the apical/basal or apical lateral. So the model energy will be lower and preferred for the apical/apical versus other pairs.
  • Lines 106-109 

    #Basal
        contact.ElementCC3D("Energy",{"Type1":"Basal","Type2":"Basal"},20)
        contact.ElementCC3D("Energy",{"Type1":"Basal","Type2":"Lateral"},20)
        contact.ElementCC3D("Energy",{"Type1":"Basal","Type2":"Lateral2"},20)​

    • Here the J terms are similar for the pairings.
  • Lines 110-113 

    #Lateral(2)
        contact.ElementCC3D("Energy",{"Type1":"Lateral","Type2":"Lateral"},4)
        contact.ElementCC3D("Energy",{"Type1":"Lateral2","Type2":"Lateral2"},4)
        contact.ElementCC3D("Energy",{"Type1":"Lateral2","Type2":"Lateral"},4)​

    • Here the J terms are similar. Compare to the other J terms. This seems to be a cell type that will associate mainly with the lateral or lateral2 (self), and are closer to the lumen (see the lumen J terms).
  • Lines 114-121 

    #Lumen
        contact.ElementCC3D("Energy",{"Type1":"Lumen","Type2":"Lumen"},0)
        contact.ElementCC3D("Energy",{"Type1":"Lumen","Type2":"Medium"},60)
        contact.ElementCC3D("Energy",{"Type1":"Lumen","Type2":"Cyto"},60)
        contact.ElementCC3D("Energy",{"Type1":"Lumen","Type2":"Apical"},15)
        contact.ElementCC3D("Energy",{"Type1":"Lumen","Type2":"Basal"},20)
        contact.ElementCC3D("Energy",{"Type1":"Lumen","Type2":"Lateral"},20)
        contact.ElementCC3D("Energy",{"Type1":"Lumen","Type2":"Lateral2"},20)​

    • Here we see J terms for cells close to the lumen Apical, lateral), versus the ones farther away( Basal, cyto). This makes sense since the kidney tubule is contacting the apical parts of the epithelial cells lining the lumen.
  • Lines 122-123 

      #Neighbor order
        contact.ElementCC3D("NeighborOrder",{},CNOrder)​

    • Per the 3.4.1 manual, this is the range over which source pixels are selected for index-copy attempts (pg 17). The default is 1. In the context of Contact Energies, it is the interaction range of the boundary energy, so a pixel on a cell boundary will interact with the pixels around it (8-neighbor) if default is 1. 
  • Lines 125-126 

    INTERNAL CONTACT ENERGIES:
        cInternal=CompuCell3DElement.ElementCC3D("Plugin",{"Name":"ContactInternal"})​

    • This creates a root object element for the "ContactInternal" plugin
  • Lines 128-141 

    #Cyto
        cInternal.ElementCC3D("Energy",{"Type1":"Cyto","Type2":"Apical"},0)
        cInternal.ElementCC3D("Energy",{"Type1":"Cyto","Type2":"Basal"},0)
        cInternal.ElementCC3D("Energy",{"Type1":"Cyto","Type2":"Lateral"},0)
        cInternal.ElementCC3D("Energy",{"Type1":"Cyto","Type2":"Lateral2"},0)
        #Apical
        cInternal.ElementCC3D("Energy",{"Type1":"Apical","Type2":"Basal"},2)
        cInternal.ElementCC3D("Energy",{"Type1":"Apical","Type2":"Lateral"},0)
        cInternal.ElementCC3D("Energy",{"Type1":"Apical","Type2":"Lateral2"},0)
        #Basal
        cInternal.ElementCC3D("Energy",{"Type1":"Basal","Type2":"Lateral"},0)
        cInternal.ElementCC3D("Energy",{"Type1":"Basal","Type2":"Lateral2"},0)
        #Lateral(2)
        cInternal.ElementCC3D("Energy",{"Type1":"Lateral","Type2":"Lateral"},0)
        cInternal.ElementCC3D("Energy",{"Type1":"Lateral2","Type2":"Lateral2"},0)​

    • For all cell-cell interactions, except for Apical-Basal, the  J term is 0. For Apical-Basal, it is 2. These J terms are for the internal cell compartments (Apical, Basal, Cyto). Which are different than the generalized cell types,
  • Lines 142-143 

     #Neighbor order
        cInternal.ElementCC3D("NeighborOrder",{},CINOrder)​

    • Assigning default neighbor order again. See Contact Energy code some paragraphs above.
  • Lines 146-150 

        pTracker=CompuCell3DElement.ElementCC3D("Plugin",{"Name":"PixelTracker"})
        boundaryPTracker=CompuCell3DElement.ElementCC3D("Plugin",{"Name":"BoundaryPixelTracker"})
        vlFlex=CompuCell3DElement.ElementCC3D("Plugin",{"Name":"VolumeLocalFlex"})
        coMass=CompuCell3DElement.ElementCC3D("Plugin",{"Name":"CenterOfMass"})
        neighborTracker=CompuCell3DElement.ElementCC3D("Plugin",{"Name":"NeighborTracker"})    ​

    • Here, other object elements are created for other plugins. The usual ones like Volume, Pixel Tracker, etc. I do not think they are used later.
  • Line 163 

     CompuCellSetup.setSimulationXMLDescription(CompuCell3DElement)​

    • This line passes the root element of the CC3DML  to the CC3D core code  for initialization (pg 57 of manual 3.7.6). So passing this one element covered all the XML thus far. This is the last line of the configureSimulation function, so later on we will call the function to get the XML and pass it into the other parts and scripts.
  • Lines 165-169 

    import CompuCellSetup
    import CompuCell
    
    sim,simThread = CompuCellSetup.getCoreSimulationObjects()
    configureSimulation(sim,Lx,Ly,Lz,Temp,Time,NOrder,CNOrder,CINOrder,debugFreq)​

    • Now we can get all the XML by calling the configureSimulation. Notice the inputs to the function.  
    • The getCoreSimulationObjects initializes the core CC3D modules. 

Q59: Cell growth on a sphere cell

content:
Hi,
are there parameters which influence the shape of a single cell in a 3D lattice beside the target volume/surface and the lambda multiplier?

Since the real volume of a sphere is almost identical with the surface out of voxels and the surface of the voxel sphere is by a factor 1.5 larger than the realistic surface, it is possible to calculate the realistic volume and surface and multiply this surface by 1.5 for the growth of a cell. With the algorithm to draw the cell it is possible to read out the volume and surface values out of CC3D. Then the factor between the real sphere surface and the voxel sphere surface can be calculated.

The algorithm to draw the cell consider the corner length of a voxel as steplength. With this approach it is checked if the center of the voxel is inside the sphere or not.

        for xr in xrange(xStart, xEnd):
            for yr in xrange(yStart, yEnd):
                for zr in xrange(zStart, zEnd):
                    rd = sqrt(
                        ((xr+(((xr+stepLength) - xr)/2.)) - x0) ** 2 +
                        ((yr+(((yr+stepLength) - yr)/2.)) - y0) ** 2 +
                        ((zr+(((zr+stepLength) - zr)/2.)) - z0) ** 2)
                    if (rd <= radiusPx):
                        steppable.cellField[xr, yr, zr] = cell​

Since this is not a perfect smooth cube and in other simulations there were more edges and peaks, I can imagine that this cell meets also the volume and surface values. 

Are there any techniques to keep the shape of the sphere during the simulation?
Is it maybe not possible because for a given volume and surface there are infinite possible 3D objects?

Thank you,
Thorsten

attachments(redirected):
.\CC3D_allanswered_linked_files\CellDiameter10.png
.\CC3D_allanswered_linked_files\CellDiameter10-MCS500.png

Answer:

content:
In addition to the volume and surface constraints the adhesion energies will also affect the sphericity of a cell. Is the system trying to not only maintain the cell's target volume and surface but also trying to avoid adhesion energy penalties? In general all of these will affect the sphericity of a cell;

• target and lambda volume
• target and lambda surface
• adhesion energies (e.g., cell to medium)
• neighbor order in pixel copy
• neighbor order in adhesion
• square vs. hexagonal lattice
• Potts temperature (which controls how quickly the surface fluctuates)

Q60: Can I manipulate or change the Gamma_S parameter (coefficient of the area term) in the Hamiltonian expression of the CoMpuCell3D Potts model?

content:

Instead of a linear constant Gamma_S, I would like to have an expression of Gamma_S term.

For example, Gamma_S = Gamma_S1*(1-Gamma_S2*GammaS1)

Is it possible in CompuCell3D? if so how to do it?

Answer:

content:
which plugin / CC3D module are you referring to? Could you post your simulation and tell us which parameter you want to modify via an arithmetic expression?

Q61: How can I restart a simulation from the last time step ?

content:
Is there a way to write a restart file so that the next simulation starts from the last time step. As for example the total MCS that I have to run is 24000. However, with the resource allocation, I can only run approximately 5000 steps continuously. How can I write a restart file so I can start from where I last ended the simulation. What needs to be changed in the number of steps ?

EDIT: How would I do it without the GUI just through the terminal? I was able to write restart file but am not sure how to specify to read from that particular directory when restarting the next stage of simulation. More specifically ,

./runScript.sh  -i  < input.cc3d> -o  < Output directory >   ...... < restart directory > ?

Answer:

content:
Have you tried following the manual?

http://pythonscriptingmanual.readthedocs.io/en/latest/restarting_simulations.html?highlight=restart

If you want to restart it make sure that you remove serialization line from the .cc3d file - you can do it as shown in the manual or simply edit .ccd file so that

<Simulation version="3.6.1">
   <XMLScript Type="XMLScript">Simulation/cellsort_2D.xml</XMLScript>
   <SerializeSimulation AllowMultipleRestartDirectories="False" FileFormat="text" OutputFrequency="100"/>
   <RestartSimulation RestartDirectory="restart"/>
</Simulation>​

<Simulation version="3.6.1">
   <XMLScript Type="XMLScript">Simulation/cellsort_2D.xml</XMLScript>
   <RestartSimulation RestartDirectory="restart"/>
</Simulation>​

attachments(redirected):
.\CC3D_allanswered_linked_files\restart_01600.zip

Comment:

content:
Thank you for the detailed description.

Answer:

content:
File attached: Run2.zip (68.93 KB)

There has been a constant error with the restart file being not found even though it is present there. I looked at the way the serializer is specified in the restart_01600 and it is the same in Run2. I am not sure why I am still getting the same error. Any help from your side would be very helpful.

Thanks and regards,
Mahua

attachments(redirected):
.\CC3D_allanswered_linked_files\Run2.zip

Answer:

content:
What error are you getting? Could you post it here?

Your restart folder is empty. But I also noticed that you are not using scipy in your simulation but still importing it. On my machine it caused problems so I commented out any references to scipy

Answer:

content:
I apologise for the empty file. Please find the attached restart file
File attached: restart_10000.zip (1.49 MB)

The error is attached here as well. It complains of the file not being there, when it is present 

[Error: IOError: loadSBMLError Wrong constructed path: RoadRunnerPy could not find /Volumes/CompuCell/Pancreatic-cancer-CC3D/Tumor_growth_simulation2/Run1/Result_atpD_032/restart_10000/ in the filesystem
Traceback (most recent call last):
File: /Users/roym/Downloads/CC3D_3.7.6/pythonSetupScripts/CompuCellPythonSimulationNewPlayer.py
Line: 351 Col: 0 execfile(CompuCellSetup.simulationPaths.simulationPythonScriptName)
execfile(CompuCellSetup.simulationPaths.simulationPythonScriptName)
File: /Volumes/CompuCell/Pancreatic-cancer-CC3D/Tumor_growth_simulation2/Run1/Result_atpD_032/restart_10000/Simulation/TumorGrowth.py
Line: 83 Col: 0 CompuCellSetup.mainLoop(sim,simthread,steppableRegistry)
CompuCellSetup.mainLoop(sim,simthread,steppableRegistry)
File: /Users/roym/Downloads/CC3D_3.7.6/pythonSetupScripts/CompuCellSetup.py
Line: 1980 Col: 0 return mainLoopNewPlayer(sim, simthread, steppableRegistry, _screenUpdateFrequency)
return mainLoopNewPlayer(sim, simthread, steppableRegistry, _screenUpdateFrequency)
File: /Users/roym/Downloads/CC3D_3.7.6/pythonSetupScripts/CompuCellSetup.py
Line: 1595 Col: 0 restartManager.loadRestartFiles()
restartManager.loadRestartFiles()
File: /Users/roym/Downloads/CC3D_3.7.6/pythonSetupScripts/RestartManager.py
Line: 340 Col: 0 self.loadSBMLSolvers()
self.loadSBMLSolvers()
File: /Users/roym/Downloads/CC3D_3.7.6/pythonSetupScripts/RestartManager.py
Line: 644 Col: 0 sbmlSolver.loadSBML(_externalPath=self.sim.getBasePath()) # this call fully initializes SBML Solver by loading sbmlMode ( relative path stored in sbmlSolver.path and root dir is passed using self.sim.getBasePath())
sbmlSolver.loadSBML(_externalPath=self.sim.getBasePath()) # this call fully initializes SBML Solver by loading sbmlMode ( relative path stored in sbmlSolver.path and root dir is passed using self.sim.getBasePath())
File: /Users/roym/Downloads/CC3D_3.7.6/pythonSetupScripts/RoadRunnerPy.py
Line: 109 Col: 0 'loadSBMLError Wrong constructed path: RoadRunnerPy could not find ' + self.absPath + ' in the filesystem')
'loadSBMLError Wrong constructed path: RoadRunnerPy could not find ' + self.absPath + ' in the filesystem')]

attachments(redirected):
.\CC3D_allanswered_linked_files\restart_10000.zip

Answer:

content:
I saw that too. and in this case it cannot find your sbml model mile (I believe in your case it might have xml extension). The serializer does not copy the sbml to the restart directory so you need to do it by hand. Simply place the sbml file in one of the restart folders you are attempting to run (in the same hierarchical location as in the original simulation i.e. inside Simulation subfolder) 

Let me know if this works

Answer:

content:
I tried copying it all locations but still the error remains the same.
'loadSBMLError Wrong constructed path: RoadRunnerPy could not find ' + self.absPath + ' in the filesystem')

IOError: loadSBMLError Wrong constructed path: RoadRunnerPy could not find /Users/roym/Downloads/restart_10000/ in the filesystem


I never received the error during the initial simulation.

Answer:

content:
so it looks like a bug. I will rerun some other simulations with SBML serializations and see what's going on. Essentially many cells that have SBML solver attached are missing the path to the sbml model. This is strange. Maybe those ar ethe new cells that got created . I will need to check that

Answer:

content:
BTW is there any chance you could try the same with CC3D 3.7.7. I know we have fixed few bugs related to copying SBML solvers during Mitosis and I wonder if this is one of those bugs that we are seeing now

Answer:

content:
Thanks so much for looking into it. I really appreciate your help. I tried with 3.7.7 version as you suggested. However, it seems to be the same error. I paste it below :


**********

File "/staging/mr2/mahuaroy/CC3D_3.7.7_RHEL/player5/CompuCellPythonSimulationCML.py", line 363, in <module>

    execfile(CompuCellSetup.simulationPaths.simulationPythonScriptName)

  File "/staging/mr2/mahuaroy/Tumor_growth_simulation2/test/restart_10000/Simulation/TumorGrowth.py", line 83, in <module>

    CompuCellSetup.mainLoop(sim,simthread,steppableRegistry)

  File "/staging/mr2/mahuaroy/CC3D_3.7.7_RHEL/pythonSetupScripts/CompuCellSetup.py", line 2072, in mainLoop

    return mainLoopCML(sim, simthread, steppableRegistry, _screenUpdateFrequency)

  File "/staging/mr2/mahuaroy/CC3D_3.7.7_RHEL/pythonSetupScripts/CompuCellSetup.py", line 1860, in mainLoopCML

    restartManager.loadRestartFiles()

  File "/staging/mr2/mahuaroy/CC3D_3.7.7_RHEL/pythonSetupScripts/RestartManager.py", line 340, in loadRestartFiles

    self.loadSBMLSolvers()

  File "/staging/mr2/mahuaroy/CC3D_3.7.7_RHEL/pythonSetupScripts/RestartManager.py", line 644, in loadSBMLSolvers

    sbmlSolver.loadSBML(_externalPath=self.sim.getBasePath()) # this call fully initializes SBML Solver by loading sbmlMode ( relative path stored in sbmlSolver.path and root dir is passed using self.sim.getBasePath())

  File "/staging/mr2/mahuaroy/CC3D_3.7.7_RHEL/pythonSetupScripts/RoadRunnerPy.py", line 109, in loadSBML

    'loadSBMLError Wrong constructed path: RoadRunnerPy could not find ' + self.absPath + ' in the filesystem')

IOError: loadSBMLError Wrong constructed path: RoadRunnerPy could not find /staging/mr2/mahuaroy/Tumor_growth_simulation2/test/restart_10000/ in the filesystem




/staging/mr2/mahuaroy/CC3D_3.7.7_RHEL/runScript.sh: line 39: 49004 Segmentation fault      ${PYTHON_EXEC} ${PREFIX_CC3D}/player5/CompuCellPythonSimulationCML.py $* --currentDir=${current_directory}

Answer:

content:
I placed the SBML (xml) file in the same location as specified in the Steppables file - in the Simulation folder.

Answer:

content:
The problem was not in the restart function but in the copySBML's function that you used in the Mitosis Steppeble (your Mitosis code was correct BTW). I fixed that along with addSBMLtoCell function and now the serialization works.

Few things though:

1. The way you initialized FPP links is not entirely correct because you end up with infinity values for lambda D after few very few MCS - you keep multiplying previous value of lambdaD by 20 each MCS. As you can tell this will led to inf values - I fixed it see lines 125-130 of your Steppables file (here is your new simulation File attached: TumorGrowth_restart.tar.gz (32.97 KB))

2. The restart file after 1000 MCS is here File attached: TumorGrowth_restart_1000.tar.gz (1.11 MB)

3. Finally to get your code working you need to patch the installation of CD3D . All you need to do is to copy the attached SBMLHelper.py (File attached: SBMLSolverHelper.py (28.46 KB)

) to <your cc3d installation directory>/pythonSetupScripts , replacing old version of SBMLHelper.py

We will be releasing new version of CC3D soon and this bug fix will be included but for now you need to do this manual step.

attachments(redirected):
.\CC3D_allanswered_linked_files\TumorGrowth_restart.tar.gz
.\CC3D_allanswered_linked_files\TumorGrowth_restart_1000.tar.gz
.\CC3D_allanswered_linked_files\SBMLSolverHelper.py

Answer:

content:
Thank so much for your time to look into the script and fix the issue. I copied the SBMLSolverHelper.py into the installation directory. Currently the restart files work. I will run on the Linux version of the CompuCell and will update you if there are any issues.

Thanks so much again.

Q62: Restart error

content:

For restarting a simulation from the last restart file written , I rename the folder and moved it to the main location where the .cc3d file is. However, this is error that is thrown out. I am not sure if there are any lines that need to be modified for the FocalPointPalsticity. Any help would be highly appreciated.

attachments(redirected):
.\CC3D_allanswered_linked_files\Screen Shot 2018-02-09 at 11.13.23 AM.png

Answer:

content:
Could you post the "toy version" of the simulation that has the problem? We could take a look at it and see where the problem is

Q63: Modeling nucleus and cytoplasm

content:
Hello. I would like to model a cell by its cytoplasm and its nucleus.  I have already tried to model each component and their interactions with internal contact and connectivity but when I tried to simulate it, I had my nucleus fragmented inside the cytoplasm, or gone out of the cytoplasm. Do you have any ideas to prevent these from happening? Thank you.

Answer:

content:
Could you post your simulation here? Cell fragmentation could be prevented by imposing surface and volume constraints. In any case if you could share your code we could try helping you out otherwise it is hard to give you specific advice without seeing all the parameters you have in your model

Answer:

content:
I guess you set J（nucleus，medium）=0，that may cause the situation you said. Try to set a large positive value to J（nucleus，medium）.

Q64: How to change a cc3d windows title?

content:
Hi,

Help needed.
Instead of the word 'graphics' (as seen in the figure), I would like have word like 'Cells' or 'Cell field' etc. How to accomplish this task? I tried to modify the python (steppable) code, but I guess it could be done also in the program itself?

Thanks in advance! 

B.r., Pauli

attachments(redirected):
.\CC3D_allanswered_linked_files\cc3d_title of the window_tikka.png

Answer:

content:
I dont think this is possible but I will add it to the TODO list

Q65: Does my simulation code crash the twedit++ when parameter scan is added for the first time?

content:
Dear reader,

I have a problem regarding crashing of twedit++ when I try to add parameter scan (see the image; question mark indicates the chrash event).
What do, how to solve? The demo cc3d codes seem to work, play, and add all the parameter scans in the world. The only solution that I can think of at the moment is to do my code again and/or reinstall the compucell3d&twedit.

Yours,
Pauli

attachments(redirected):
.\CC3D_allanswered_linked_files\parameter scan chrashes the simulation_why what to do.png

Answer:

content:
Hi,

I changed the folder of my simulation to compucell...\Book_ChapterDemos etc. , and it I got the scan working! Splendid! I wonder how I can give rights to other folders for cc3d sim runs? 

B.r.,
Pauli
me
Europe

Answer:

content:
What platform are you running on? Windows, Mac, linux, ...? It sounds like the parameter scan task is trying to write to a directory you didn't have access to.

I've also noticed crashes when you try to save the file after defining the parameter scans. In my case though the files are written correctly and can be re-loaded into twedit, or run from the command line.

Comment:

content:
Windows. I use the same directory where my compucell3d is. I have still experienced problems in starting to run the simulation code in my new directory inside this folder. The program does not crash, but it does not play the simulation code. I have attached the run file.

processCommandLineOptions




opts= []
args= []
FILE LIST= []
pixmap= <PyQt5.QtGui.QPixmap object at 0x0000000005586278>
INIT SYNC self.modifiedPluginsDataStringList= []
pluginModules= ['PluginCCDCPPHelper', 'PluginCCDMLHelper', 'PluginCCDProject', 'PluginCCDPythonHelper', 'PluginCompuCell3D']
plugin name= PluginCCDCPPHelper



 ##################loading source for  C:\CompuCell3D-64bit\Twedit++5\Plugins\PluginCCDCPPHelper.py
plugin name= PluginCCDMLHelper



 ##################loading source for  C:\CompuCell3D-64bit\Twedit++5\Plugins\PluginCCDMLHelper.py
['C:\\CompuCell3D-64bit\\lib\\PythonDeps', 'C:\\CompuCell3D-64bit\\Twedit++5', 'C:\\CompuCell3D-64bit\\Twedit++5\\Plugins', 'C:\\CompuCell3D-64bit\\Python27\\python27.zip', 'C:\\CompuCell3D-64bit\\Python27\\DLLs', 'C:\\CompuCell3D-64bit\\Python27\\lib', 'C:\\CompuCell3D-64bit\\Python27\\lib\\plat-win', 'C:\\CompuCell3D-64bit\\Python27\\lib\\lib-tk', 'C:\\CompuCell3D-64bit\\Python27', 'C:\\CompuCell3D-64bit\\Python27\\lib\\site-packages', 'C:\\CompuCell3D-64bit\\Python27\\lib\\site-packages\\win32', 'C:\\CompuCell3D-64bit\\Python27\\lib\\site-packages\\win32\\lib', 'C:\\CompuCell3D-64bit\\Python27\\lib\\site-packages\\Pythonwin', 'C:\\CompuCell3D-64bit\\lib\\python', 'C:\\CompuCell3D-64bit\\pythonSetupScripts']
plugin name= PluginCCDProject



 ##################loading source for  C:\CompuCell3D-64bit\Twedit++5\Plugins\PluginCCDProject.py
plugin name= PluginCCDPythonHelper



 ##################loading source for  C:\CompuCell3D-64bit\Twedit++5\Plugins\PluginCCDPythonHelper.py
plugin name= PluginCompuCell3D



 ##################loading source for  C:\CompuCell3D-64bit\Twedit++5\Plugins\PluginCompuCell3D.py
************PLUGIN NAME= PluginCompuCell3D
name=PluginCompuCell3D
fileName=C:\CompuCell3D-64bit\Twedit++5\Plugins\PluginCompuCell3D.py
author=Maciej Swat
autoactivate=True
deactivateable=False
version=0.9.0
className=CC3DApp
packageName=__core__
shortDescription=PLugin linking Twedit++5 with CompuCell3D
longDescription=This plugin provides functionality to link Twedit with CompuCell3D

LOADING  PluginCompuCell3D from  C:\CompuCell3D-64bit\Twedit++5\Plugins\PluginCompuCell3D.py
loading source for  C:\CompuCell3D-64bit\Twedit++5\Plugins\PluginCompuCell3D.py
version= 0.9.0  className= CC3DApp  pluginClass= <class 'PluginCompuCell3D.CC3DApp'>
PORT= -1
__initActions
CC3D CONSTRUCTOR
WILL TRY TO ACTIVATE  <PluginCompuCell3D.CC3DApp object at 0x0000000005921678>
ACTIVATED
************PLUGIN NAME= PluginCCDProject
name=PluginCCDProject
fileName=C:\CompuCell3D-64bit\Twedit++5\Plugins\PluginCCDProject.py
author=Maciej Swat
autoactivate=True
deactivateable=False
version=0.9.0
className=CC3DProject
packageName=__core__
shortDescription=Plugin to manage CC3D Projects
longDescription=This plugin provides functionality that allows users to manage *.cc3d projects

LOADING  PluginCCDProject from  C:\CompuCell3D-64bit\Twedit++5\Plugins\PluginCCDProject.py
loading source for  C:\CompuCell3D-64bit\Twedit++5\Plugins\PluginCCDProject.py
version= 0.9.0  className= CC3DProject  pluginClass= <class 'PluginCCDProject.CC3DProject'>
libpng warning: iCCP: known incorrect sRGB profile
libpng warning: iCCP: known incorrect sRGB profile
libpng warning: iCCP: known incorrect sRGB profile
WILL TRY TO ACTIVATE  <PluginCCDProject.CC3DProject object at 0x0000000005921D38>
ACTIVATED
************PLUGIN NAME= PluginCCDPythonHelper
name=PluginCCDPythonHelper
fileName=C:\CompuCell3D-64bit\Twedit++5\Plugins\PluginCCDPythonHelper.py
author=Maciej Swat
autoactivate=True
deactivateable=True
version=0.9.0
className=CC3DPythonHelper
packageName=__core__
shortDescription=Plugin which assists with CC3D Python scripting
longDescription=This plugin provides provides users with CC3D Python code snippets - making Python scripting in CC3D more convenient.

LOADING  PluginCCDPythonHelper from  C:\CompuCell3D-64bit\Twedit++5\Plugins\PluginCCDPythonHelper.py
loading source for  C:\CompuCell3D-64bit\Twedit++5\Plugins\PluginCCDPythonHelper.py
version= 0.9.0  className= CC3DPythonHelper  pluginClass= <class 'PluginCCDPythonHelper.CC3DPythonHelper'>
WILL TRY TO ACTIVATE  <PluginCCDPythonHelper.CC3DPythonHelper object at 0x0000000005921438>
menuName=  Adhesion Flex
menuName=  Cell Attributes
menuName=  Cell Constraints
menuName=  Cell Manipulation
menuName=  Chemical Field Manipulation
menuName=  Chemotaxis
menuName=  Distances, Vectors, Transformations
menuName=  Elasticity
menuName=  Extra Fields
menuName=  Extra Fields Automatic Tracking
menuName=  Focal Point Placticity
menuName=  Inertia Tensor
menuName=  Mitosis
menuName=  Parameter Scan Command Line
menuName=  Python Utilities
menuName=  SBML Solver
menuName=  Scientific Plots
menuName=  Scientific Plots Histograms
menuName=  Secretion / Uptake
menuName=  Simulation
menuName=  Visit
************PLUGIN NAME= PluginCCDCPPHelper
name=PluginCCDCPPHelper
fileName=C:\CompuCell3D-64bit\Twedit++5\Plugins\PluginCCDCPPHelper.py
author=Maciej Swat
autoactivate=True
deactivateable=True
version=0.9.0
className=CC3DCPPHelper
packageName=__core__
shortDescription=Plugin assists with CC3D C++ module development scripting
longDescription=This plugin provides provides users with CC3D C++ code generator and code snippets - making CC3D C++ plugin and steppable development  more convenient.

LOADING  PluginCCDCPPHelper from  C:\CompuCell3D-64bit\Twedit++5\Plugins\PluginCCDCPPHelper.py
loading source for  C:\CompuCell3D-64bit\Twedit++5\Plugins\PluginCCDCPPHelper.py
version= 0.9.0  className= CC3DCPPHelper  pluginClass= <class 'PluginCCDCPPHelper.CC3DCPPHelper'>
WILL TRY TO ACTIVATE  <PluginCCDCPPHelper.CC3DCPPHelper object at 0x000000000588B558>
menuName=  Cell Attributes
menuName=  Include
menuName=  Module Setup
menuName=  Visit
menuName=  XML Utils
ACTIVATED
************PLUGIN NAME= PluginCCDMLHelper
name=PluginCCDMLHelper
fileName=C:\CompuCell3D-64bit\Twedit++5\Plugins\PluginCCDMLHelper.py
author=Maciej Swat
autoactivate=True
deactivateable=True
version=0.9.0
className=CC3DMLHelper
packageName=__core__
shortDescription=Plugin which assists with CC3D Python scripting
longDescription=This plugin provides provides users with CC3D Python code snippets - making Python scripting in CC3D more convenient.

LOADING  PluginCCDMLHelper from  C:\CompuCell3D-64bit\Twedit++5\Plugins\PluginCCDMLHelper.py
loading source for  C:\CompuCell3D-64bit\Twedit++5\Plugins\PluginCCDMLHelper.py
version= 0.9.0  className= CC3DMLHelper  pluginClass= <class 'PluginCCDMLHelper.CC3DMLHelper'>
WILL TRY TO ACTIVATE  <PluginCCDMLHelper.CC3DMLHelper object at 0x000000000588B798>
ACTIVATED
_settingName= RecentProjects
recentItems= [u'C:\\CompuCell3D-64bit\\pauli\\aggregation\\PKD_paper_codes\\isoCyst\\isoCyst.cc3d', u'C:\\CompuCell3D-64bit\\Demos\\BookChapterDemos_ComputationalMethodsInCellBiology\\VascularTumor\\VascularTumor.cc3d', u'C:\\CompuCell3D-64bit\\pauli\\aggregation\\PKD_paper_codes\\tubule\\PKD.cc3d', u'C:\\CompuCell3D-64bit\\pauli\\movements\\m19\\okok\\peitto.cc3d', u'C:\\CompuCell3D-64bit\\pauli\\aggregation\\PTAtoRV-STANDARD\\aggregration\\aggregration.cc3d', u'C:\\CompuCell3D-64bit\\Demos\\BookChapterDemos_ComputationalMethodsInCellBiology\\okok\\peitto.cc3d', u'C:\\CompuCell3D-64bit\\Demos\\BookChapterDemos_ComputationalMethodsInCellBiology\\cellsorting2\\cellsorting.cc3d', u'C:\\CompuCell3D-64bit\\Demos\\BookChapterDemos_ComputationalMethodsInCellBiology\\cellsorting\\cellsorting.cc3d']
__loadRecentProject
GOT SUSTOM SETTINGS :  C:\CompuCell3D-64bit\Demos\BookChapterDemos_ComputationalMethodsInCellBiology\okok\Simulation\_settings.sqlite
projParent= peitto.cc3d
__openXMLPythonInEditor pdh.cc3dSimulationData.xmlScript= C:\CompuCell3D-64bit\Demos\BookChapterDemos_ComputationalMethodsInCellBiology\okok\Simulation\peitto.xml
__openXMLPythonInEditor pdh.cc3dSimulationData.xmlScriptResource.path= C:\CompuCell3D-64bit\Demos\BookChapterDemos_ComputationalMethodsInCellBiology\okok\Simulation\peitto.xml
opening file  C:\CompuCell3D-64bit\Demos\BookChapterDemos_ComputationalMethodsInCellBiology\okok\Simulation\peitto.xml
DONE READING: file  C:\CompuCell3D-64bit\Demos\BookChapterDemos_ComputationalMethodsInCellBiology\okok\Simulation\peitto.xml
opening file  C:\CompuCell3D-64bit\Demos\BookChapterDemos_ComputationalMethodsInCellBiology\okok\Simulation\peitto.py
DONE READING: file  C:\CompuCell3D-64bit\Demos\BookChapterDemos_ComputationalMethodsInCellBiology\okok\Simulation\peitto.py
opening file  C:\CompuCell3D-64bit\Demos\BookChapterDemos_ComputationalMethodsInCellBiology\okok\Simulation\peittoSteppables.py
DONE READING: file  C:\CompuCell3D-64bit\Demos\BookChapterDemos_ComputationalMethodsInCellBiology\okok\Simulation\peittoSteppables.py
projParent= peitto.cc3d



 RESOURCENAME
projParent= peitto.cc3d
CurrentItem= peitto.cc3d  parent= None
getFullPath= projParent= peitto.cc3d

projectFullPath= C:\CompuCell3D-64bit\Demos\BookChapterDemos_ComputationalMethodsInCellBiology\okok\peitto.cc3d
TRY TO FIGURE OUT PORT






CHECKING PORT= 47406
CHECKING PORT= 47407
established empty port= 47407
self.cc3dPath= C:\CompuCell3D-64bit\compucell3d.bat
Executing Popen command with following arguments= ['C:\\CompuCell3D-64bit\\compucell3d.bat', '--port=47407', '-i', u'C:\\CompuCell3D-64bit\\Demos\\BookChapterDemos_ComputationalMethodsInCellBiology\\okok\\peitto.cc3d']
starting CC3D
['C:\\CompuCell3D-64bit\\player5', 'C:\\CompuCell3D-64bit\\lib\\python', 'C:\\CompuCell3D-64bit\\Python27\\python27.zip', 'C:\\CompuCell3D-64bit\\Python27\\DLLs', 'C:\\CompuCell3D-64bit\\Python27\\lib', 'C:\\CompuCell3D-64bit\\Python27\\lib\\plat-win', 'C:\\CompuCell3D-64bit\\Python27\\lib\\lib-tk', 'C:\\CompuCell3D-64bit\\Python27', 'C:\\CompuCell3D-64bit\\Python27\\lib\\site-packages', 'C:\\CompuCell3D-64bit\\Python27\\lib\\site-packages\\win32', 'C:\\CompuCell3D-64bit\\Python27\\lib\\site-packages\\win32\\lib', 'C:\\CompuCell3D-64bit\\Python27\\lib\\site-packages\\Pythonwin']
compucell3d.pyw:   type(argv)= <type 'list'>
compucell3d.pyw:   argv= ['--port=47407', '-i', 'C:\\CompuCell3D-64bit\\Demos\\BookChapterDemos_ComputationalMethodsInCellBiology\\okok\\peitto.cc3d', '--currentDir=C:\\CompuCell3D-64bit']
CONSOLE PARENT= <UI.UserInterface.DockWidget object at 0x000000000AF34678>
self.baseFont.fixedPitch()= True
TWEDIT socket.socketDescriptor()= <sip.voidptr object at 0x000000000AF2E850>
TRY TO FIGURE OUT PORT






CHECKING PORT= 47406
CHECKING PORT= 47407
CHECKING PORT= 47408
established empty port= 47408

CALL establishConnection
self.__fileName= C:\CompuCell3D-64bit\Demos\BookChapterDemos_ComputationalMethodsInCellBiology\okok\peitto.cc3d
GOT SUSTOM SETTINGS :  C:\CompuCell3D-64bit\Demos\BookChapterDemos_ComputationalMethodsInCellBiology\okok\Simulation\_settings.sqlite
currentIteration= [3]
multiplicativeFactors= [1]



 iterationId= 3
Error Parsing Parameter scan file



Traceback (most recent call last):
  File "C:\CompuCell3D-64bit\player5\compucell3d.pyw", line 258, in <module>
    error_code = main(sys.argv[1:])
  File "C:\CompuCell3D-64bit\player5\compucell3d.pyw", line 239, in main
    mainWindow.viewmanager.processCommandLineOptions(cml_args)
  File "C:\CompuCell3D-64bit\player5\Plugins\ViewManagerPlugins\SimpleTabView.py", line 687, in processCommandLineOptions
    self.__runSim()
  File "C:\CompuCell3D-64bit\player5\Plugins\ViewManagerPlugins\SimpleTabView.py", line 1815, in __runSim
    self.prepareSimulation()
  File "C:\CompuCell3D-64bit\player5\Plugins\ViewManagerPlugins\SimpleTabView.py", line 1775, in prepareSimulation
    self.__loadSim(file)
  File "C:\CompuCell3D-64bit\player5\Plugins\ViewManagerPlugins\SimpleTabView.py", line 1015, in __loadSim
    self.__loadCC3DFile(fileName)
  File "C:\CompuCell3D-64bit\player5\Plugins\ViewManagerPlugins\SimpleTabView.py", line 1169, in __loadCC3DFile
    psu.replaceValuesInSimulationFiles(_pScanFileName=pScanFilePath, _simulationDir=customOutputPath)
  File "C:\CompuCell3D-64bit\pythonSetupScripts\ParameterScanUtils.py", line 757, in replaceValuesInSimulationFiles
    self.replaceValuesInXMLFile(_parameterScanDataMap = parameterScanDataMap , _xmlFile=fullFilePath)
  File "C:\CompuCell3D-64bit\pythonSetupScripts\ParameterScanUtils.py", line 674, in replaceValuesInXMLFile
    elem=self.getXMLElementFromAccessPath(root_element,accessPathList)
  File "C:\CompuCell3D-64bit\pythonSetupScripts\ParameterScanUtils.py", line 629, in getXMLElementFromAccessPath
    tmpElement=tmpElement.getFirstElement(arg[0],attrDict) if attrDict is not None else tmpElement.getFirstElement(arg[0])
AttributeError: 'NoneType' object has no attribute 'getFirstElement'
CLOSING LOCAL SOCKET
"SIMULATION FINISHED"
​

Answer:

content:
Could you post s short version of your simulation where the problem occurs? 

Answer:

content:
Hi,

Here is a short video about the crash (that occurs around 56s in the video) when I start to add the parameter scan to my cc3d model in twedit++5.

So the player goes on, but the twedit++5 crashes after attempting to add the scan.

Best,
Pauli

File attached: adding parameter scan results to crash.mp4 (2.12 MB)

attachments(redirected):
.\CC3D_allanswered_linked_files\adding parameter scan results to crash.mp4

Answer:

content:



What to do, here is what I also get?:

attachments(redirected):
.\CC3D_allanswered_linked_files\cpp_error.png

Answer:

content:
The path to the .piff file is odd, there is a single slash whereas all the other directory delimiters are back slashes. Check your code for a bad path to the piff file. Also, take a look in the the .cc3d file and see if the piff file is listed there and make sure the path has the correct delimiters. It looks like you are on a Windows system and Windows is "supposed" to be able to handle slashes or backslashes in paths, but that could still be the problem.

If you are using a .piff file you can include it in the .cc3d file by importing into the project in Twedit++. Use "Add resource", tell it the file is a piff, then browse to the file. If the file is already in the Simulation directory Twedit++ will complain that it can't copy the file. Regardless, you just need to make sure there is a copy in the Simulation folder and that that is included in the project's .cc3d file. That way when the various files are all copied for the parameter scan the .piff file goes with them. So, in general, the .piff file should be in the Simulation folder for the CC3D project.

Answer:

content:
Hi,

Thank you for your answer!
I did what you recommended (see peitto.cc3d), and inserted the piff file to .cc3d file separately.
I encountered some problems (see below), but was able to add the parameter scan, and modify it manually line-by-line. So I now have parameter scan, but If I want to add new elements to scan, I need to write them directly to the xml code myself.

Next question would be, how to save the result file in the parameter_scan subfolder (and not a folder, so that the previous results would not wipe the one iteration round older file/s)?

E.g. what to do in steppables.py file to lines like(?):
FileName = "C:/pyintro/positions.csv" # you can add some specific facts to the file if needed
self.File = open(FileName,"w",0) # the 0 is for initial memory 'flush', needed for total print of the file
self.File.write("cell_no,time_mcs,x_position_(px),y_position_(px),z_position_(px)\n")

Best,
Pauli

File attached: peitto.cc3d (407 Bytes)
File attached: errorlog_param_tikka16318.txt (3.22 KB) 

attachments(redirected):
.\CC3D_allanswered_linked_files\cpp_error2.png

Comment:

content:
That is the message the parameter scan process gives when the scan is complete. If you have old folders in the directory structure then when the scan tries to create the folder for a new scan it fails. The scan process can't differentiate between old and unneeded scan folders and new folders created during the current scan process.

You just need to delete all those old numbered folders before restarting a scan.

If you restart a scan you also have to edit the scan xml file and set the pointers back to zero.
For example, from one of my scan files;
<Parameter CurrentIteration="3" Name="pbpk_kGutabs" Type="PYTHON_GLOBAL" ValueType="float">
CurrentIteration needs to be reset back to zero.

Comment:

content:
Hi,
Thanks! I got it now.
In addition, I can save the different iteration files under different names to the same scan folder.
Regards,
Pauli

Comment:

content:
Ah no!! The crashing problem reoccurred, but in a different place and context. I still wanted to scan parameters automatically via scan editor, but the twetedit program crashes immediately when I push ok after inserting the values in the 'scannable parameters' window. What to do, continue with manual approach or rewrite the whole code in other folder etc.?

Comment:

content:
Oo! I think I now got it all finally. I changed the folder, and deleted all extra files from simulation folder (+chekced the .cc3d file again), and now I can do the scans normally with the editor. Thanks!

Answer:

content:
The diffusion constant scan was not working properly, here it how it should be modified, please correct it for the automatic version!
The first scan is ok:

<ParameterScan version="3.7.0">
<OutputDirectory>NewSimulation_ParameterScan</OutputDirectory>
<ParameterList Resource="Simulation/NewSimulation.xml">
<Parameter CurrentIteration="0" Name="GlobalDiffusionConstant" Type="XML_CDATA" ValueType="float">
[['CompuCell3D','Revision','20180621','Version','3.7.8'],['Steppable','Type','DiffusionSolverFE'],['DiffusionField','Name','Wnt9b'],['GlobalDiffusionConstant']]
<Values>0,1,3,4,6</Values> 
</Parameter>
<!-- <Parameter CurrentIteration="0" Name="GlobalDiffusionConstant" Type="XML_CDATA" ValueType="float"> -->
<!-- [['CompuCell3D','Revision','20180621','Version','3.7.8'],['Steppable','Type','DiffusionSolverFE'],['DiffusionField','Name','Wnt9b'],['DiffusionData'],['GlobalDiffusionConstant']] --> WRONG! do not know why, but does not work with diffusion data field
<!-- <Values>0.5,1.625,2.75,3.875,5.0</Values> -->
<!-- </Parameter> -->
</ParameterList>
</ParameterScan>

Comment:

content:
oh no, I checked the param scan xml files, and it had not changed the D values... aaa

Answer:

content:
I reinstalled the newest compucell3d (21618 version), and the problem did not disappear with a test simulation.


Good things:
The cellsorting param demo works (but it does not have D or gamma values), 
and also the test simulation, if I modify the param scan specification code excluding diffusion data -field:

<ParameterScan version="3.7.0">
...
      <Parameter CurrentIteration="0" Name="GlobalDiffusionConstant" Type="XML_CDATA" ValueType="float">
         [['CompuCell3D','Revision','20180621','Version','3.7.8'],['Steppable','Type','DiffusionSolverFE'],['DiffusionField','Name','wnt'], (???)['GlobalDiffusionConstant']]
         <Values>...

However, then the (D) values are always the ones at the beginning, e.g. 1,1,...x amount of scans
Have you tested that parameter scan for diffusionsolverFE/DE etc. elements work in the latest cc3d?
You do not necessarily need to drive my simulation, any would do.

Thanks.

Answer:

content:
I am attaching ParameterScanUtils.py file
File attached: ParameterScanUtils.py (31.01 KB)

attachments(redirected):
.\CC3D_allanswered_linked_files\ParameterScanUtils.py

Answer:

content:
I tried Maciek's fix.

A PScan of GlobalDiffusionConstant in C:\CompuCell3D-64bit\Demos\SteppableDemos\DiffusionSolverFE\DiffusionSolverFE now works. (It didn't before)

I also tried Pauli's simulation and it also works now. I even scanned two of the diffusion parameters simultaneously.

A minor annoyance. Before doing a pscan you should open CC3D separately from Twedit++ and set the [Tools][Configuration][Project location] and [output] directories correctly for your project. Otherwise CC3D is likely to try to write pscan output to some other directory, which may already have pscan output.

Q66: How does the external potential plugin work

content:
Hi, 
I have been working with the ExternalPotential plugin. I couldn't find the model equations for this plugin in the manuals. After going through the source code, I came to the conclusion that this could be the underlying model equation,
  $\Delta H_{force}=\lambda\sum_{pixels}\left(\vec{i'}-\vec{i}\right)$ΔH_ƒ orce=λ∑_pixels(→i'−→i)   for the pixelbased calculations and   $\Delta H=\lambda\left(\vec{i_{com}'}+\vec{i_{com}}\right)$ΔH=λ(→i_com'+→i_com) for the center of mass based calculations.

Is my understanding correct? What changes should I make to the equations? Why is the center of mass based equation additive in terms of energy?

Thanks.

Answer:

content:
Yes, your understanding is close to being correct. For the COM based algorithm it computes a displacement vector that would result if the pixel copy is accepted and takes a scalar product of this vector with the vector whose components in the in XML are  specified using:

<ExternalPotentialParameters CellType="NonCondensing" x="-10" y="0" z="0"/>​

Q67: Cyst formation paper (Belmonte 2016) reproduction problems for reuse in my model

content:
Dear Reader,

I try to reproduce your PKD model code in my modelling process (pretubular aggregate to renal vesicle, including MET) by changing the parameters so that the model output similar as I do have. However, I cannot open the codes (from http://compucell3d.org/Models @PKD_paper_codes) in my compucell3d (see attachement).


For instance, transferring these files (including *.ccds) to my computer's (cc3d) modelling file and making new simulation files.

Thanks for all possible help.​

Best regards,
Pauli



File attached: isoCyst_error.txt (434.04 KB)

attachments(redirected):
.\CC3D_allanswered_linked_files\isoCyst_error.txt

Answer:

content:
Looking at your log file it looks like you had a succesfull run followed by a different (?) run that failed with the error;

self.errorConsole.playerMainWidget= <Plugins.ViewManagerPlugins.SimpleTabView.SimpleTabView object at 0x000000000756DEE8>​

Q68: Makefile Problem in CompuCelle3D 3.7.7 Installation

content:
I am in the process of trying to install CompuCell 3D 3.7.7 on a machine that runs Linux Mint 18.

I set up the makefile according to the instructions provided, but when I tried to do the compile step, I got the following error:


core/pyinterface/CC3DXML/CMakeFiles/_CC3DXML.dir/flags.make:10: *** missing separator. Stop.
CMakeFiles/Makefile2:5407: recipe for target 'core/pyinterface/CC3DXML/CMakeFiles/_CC3DXML.dir/all' failed
make[1]: *** [core/pyinterface/CC3DXML/CMakeFiles/_CC3DXML.dir/all] Error 2
Makefile:127: recipe for target 'all' failed
make: *** [all] Error 2

I would greatly appreciate if anyone is able to advise me on how to address this problem.

Answer:

content:
The problem you have mentioned is very generic. there many libraries that CompuCell3D depends on it is possible that there is some library mismatch. Could you please download try  pre-complied linux package available for CompuCell3D from here. Please let me know whether it worked for you or not.

Answer:

content:
Additional thing that helps us debug is the output of the make command that is executed as follows:

make VERBOSE=1​

Q69: How to create a cell sheet consisting of hexagonal cells?

content:
Dear all,
I am trying to create a hexagonal cell sheet, as shown below. Each cell is hexagonal and made up of hexagonal pixels. What should I do?
Many thanks.

cell sheet:

each cell:

attachments(redirected):
.\CC3D_allanswered_linked_files\QQ截图20180228170631.png
.\CC3D_allanswered_linked_files\QQ截图20180228165632.png

Answer:

content:
You can specify that CC3D should use a hexagonal lattice with;
  <LatticeType>Hexagonal</LatticeType>
in the Potts section of the project's XML file.

Also, take a look at the hexagonal lattice demo in;
   C:\CompuCell3D\Demos\SimulationSettings\HexLattice

Finally, for the mathematical details of a hexagonal lattice see;
   http://www.compucell3d.org/BinDoc/cc3d_binaries/Manuals/HexagonalLattice.pdf

Comment:

content:
What should I do if I want to make each cell hexagonal using hexagonal lattice? As shown in the second picture.

Answer:

content:
I've created an extended example of hexagonal cells on a hexagonal lattice.
Here is a zip file of the CC3D code -- File attached: Hexcells2b.zip (7.91 KB)
This uses the blob initializer, which doesn't know about the hex lattice so it will take 10K or more MCS to get the cells to relax into hexagons.

Note that if you want the cells to move then they can not always be perfect hexagons. If you want perfect hexagons then the cells will have to be frozen and wont be able to move.

attachments(redirected):
.\CC3D_allanswered_linked_files\screenshot.png
.\CC3D_allanswered_linked_files\Hexcells2b.zip

Q70: How to degrade ECM pixels?

content:
Dear all,
In my simulation once the cancer cell comes in to contact with ECM,  starts to secrete MMP and once the MMP concentration in ECM lattice sites reach a threshold, they switches their type from ECM to medium or they die. but my code doesnt work.(ECM is frozen)
Sincerely,

attachments(redirected):
.\CC3D_allanswered_linked_files\Capture.png

Answer:

content:
You can't set the volume and surface qualities for MEDIUM or for a single pixel. You should be able to just set the cell type of the pixel to MEDIUM (cell.type=0).

So try just
   px.type=0
and remove the px.target... and px.lambda... commands

If that simple approach doesn't work try
    self.cellField.set(px,CompuCell.getMediumCell())

Comment:

content:
Yeah target volume was just an assupmtion and was uncommented and in fact i had used px.type=0 but it didnt work. thank you  very much for your help, self.cellField.set(px,CompuCell.getMediumCell()) solved the issue.

Answer:

content:
We have implemented the MMP-dependent ECM degradation in the following paper:

https://www.nature.com/articles/srep19905

If you want to do something like this, I can share the code.

Answer:

content:
Sandeep that is very nice work. If you don't mind sharing the code we would like to add it to the models listed at http://www.compucell3d.org/Models. You can email the files to me at jsluka@indiana.edu.

thanks!

Answer:

content:
@James: Sent the CC3D files to your email. 

Sandeep

Q71: Celldraw in CompuCell3D 3.7.7

content:
I am running CompuCell3D 3.7.7 on Linux Mint 17.1. I followed the instructions to compile and install from the source, and for the project I am doing, I need to use Celldraw. However, the celldraw.sh file is missing. Is there some way that I can get it, or if not, is there another way that I can run Celldraw?

Answer:

content:
Did you check the previous versions of CC3D (e.g. 3.6.x) to see if they have the celldraw.sh file? There shouldn't be any difference in celldraw across the last couple of CC3D versions.

If that doesn't work, I have a rough python/CC3D script that will read an image file and convert it to a cc3d cell field. The image file is typically just created in something like MSPaint. If you need something like that let me know.

Answer:

content:
We have changed the Qt version and as a result we do not have Cell Draw in CC3D 3.7.7 because this part of code was not ported but as Jim mentioned if you install 3.7.5 you will get CellDraw. So what you need to do is to install PyQt4 and then you do not even have to compile CC3D but rather CellDraw folder to a directory and run directly from there. If you need help with this let us know and we can assist you

Answer:

content:
Here are instructions on how to run CellDraw with newer versions of CC3D

http://www.compucell3d.org/SrcBin/CellDrawPyQt4

Comment: