Dependencies:
- A C++ compiler that supports C++11
- Boost libraries: system, filesystem, iostreams, random
- CMake 2.6 or newer
- Gzip (optional)
On Ubuntu 16.04 all dependencies can be installed with the following command:
sudo apt-get install cmake libboost-filesystem-dev libboost-iostreams-dev libboost-random-dev gcc g++ python gzip```
Build:
```bash
mkdir build
cd build
cmake ../
make
To run a simulation you need to call the binary with an xml file:
cd examples
mkdir out
../bin/simulator iteratedgrowth_csc_otl.xmlNote that the code will crash if the output folder (<OutPath>out/</OutPath>) does not exist!
The simulation settings are stored in a single xml file containing 3 nodes: Simulation, Solver, and Experiment. An example xml can be found in the examples folder.
The model node specifies the growth model, which includes the species and transitions, and the related settings. The growth model is specified via the type attribute, and may be:
- Simple: constant growth with one cell type without cell death
- CSC: stem cell driven growth with a specified number of division for differentiated cells, and all cells may go into apoptosis.
| Element | Description |
|---|---|
| DivisionRate | Number of times cells divide per day |
| DivisionRateSD | Standard deviation of the division rates |
| DivisionRateLogNormal | Use lognormal instead of normal distribution for division rates |
| DeathRate | Death rate of the cells |
| Seed | Seed for generating variable division rates |
| Element | Description |
|---|---|
| MaxAgeTransitCell | Number of transitions for DCs |
| DivisionRateStemCell | Number of times CSCs divide per day |
| DivisionRateTransitCell | Number of times DCs divide per day |
| DeathRateStemCell | Number of times CSCs die per day |
| DeathRateTransitCell | Number of times DCs die per day |
| ProbSymmStemCellDivision | Probability of a symmetric CSC division |
| ProbSymmStemCellDifferentiation | Probability of a symmetric CSC differentiation |
| DivisionRateSD | Standard deviation of the division rates |
| Seed | Seed for generating variable division rates |
| InitialFractions | Comma-separated list of MaxAgeTransitCell+1 initial fractions |
| InitialStemCellFraction | Initial fraction of CSCs, all other species get identical initial fractions |
The solver node specifies the solver and related settings. The solver is set via the type attribute, which, for now, can only be GillspieOTL, a solver using ordinary tau leaping gillespie. The only setting for GillespieOTL solver is tau, which refers to the length (in days) of one tau-step.
The experiment node specifies the kind of experiments and the experiment specific settings. The experiment is specified via the type attribute, and can be:
- ConstantGrowth: cells grow for a given time
- IteratedGrowth: cells grow untill a target number is reached and then a subset is passed to the next generation.
| Element | Description |
|---|---|
| SimulationTime | Simulation time in days |
| Seed | Seed used for stochastic growth and passage |
| InitSeed | Seed used for the initialization (if ommitted, Seed is used) |
| InitialPopulationSize | Initial number of cells |
| InitFile | File with initial clone distribution |
| InitUniform | Number of clones over which the initial cells are uniformly distributed |
| Name | Simulation name, used for filenames of the output |
| SaveFreq | Frequency for generating output |
| Outpath | Path to store output to |
| SaveXML | Export all settings (including defaults) in an xml file and save it to Outpath |
| gzip | Gzip all output files |
| UseSimDir | Create a folder in Outpath that contains all generated output |
| Element | Description |
|---|---|
| CriticalPopulationSize | Population size after which a growth step is stopped |
| NumberOfCellsToKeep | Number of cells to keep after passage |
| NumberOfPassages | Number of growth and passage cycles |
| MaxPassTime | Maximum time a growth step may take |
| StopSimIfNPassNotReached | Stop the simulation if MaxPassTime is reached |
| Seed | Seed used for stochastic growth and passage |
| InitSeed | Seed used for the initialization (if ommitted, Seed is used) |
| InitialPopulationSize | Initial number of cells |
| InitFile | File with initial clone distribution |
| InitUniform | Number of clones over which the initial cells are uniformly distributed |
| Name | Simulation name, used for filenames of the output |
| SaveFreq | Frequency for generating output |
| Outpath | Path to store output to |
| SaveXML | Export all settings (including defaults) in an xml file and save it to Outpath |
| gzip | Gzip all output files |
| UseSimDir | Create a folder in Outpath that contains all generated output |
Alternatively, you can use Docker to build the software:
docker build -t ssa .
Then, you can run the software with:
cd examples/
mkdir out
docker run -u $(id -u):$(id -g) -v $PWD:/data/ ssa /data/iteratedgrowth_csc_otl_docker.xml
Note that this command mounts the current directory as /data/ inside the container. Therefore, OutPath in the xml must start with /data/ as well.
We thank the developers of the TinyXML-2 library for providing this library (https://github.com/leethomason/tinyxml2).