PYSB Simulation Framework

A framework for simulating MAPK pathway dynamics using PySB.

Setup

1. Clone this repository:

git clone <repository-url>
cd <repository-name>

2. Make the setup script executable and run it:

chmod +x setup_pysb_env.sh
./setup_pysb_env.sh

3. Activate the environment:

source ~/.bashrc
conda activate pysb_env

Model Structure

The model consists of several components (see src/main_my.py, startLine: 20, endLine: 39):

• MAPK pathway components
• AP1 transcription factors
• Cell state transitions
• Oscillatory behavior

Running Simulations

Single Cell Simulation

For deterministic ODE simulations (see src/main.py, startLine: 23, endLine: 170):

python src/main.py \
    --cell-line mutant \
    --drug-concentration 1.0 0.5 \
    --output results/single_cell.h5 \
    --plot-output results/single_cell.png

Population Simulation

For stochastic simulations (see src/main_my.py, startLine: 128, endLine: 162):

python src/main_my.py \
    --cell-line mutant \
    --drug-concentration 1.0 0.5 \
    --output results/population.h5 \
    --plot-output results/population.png \
    --n-cells-plot 5

Command Line Arguments

• --cell-line: Choose between 'mutant' or 'wildtype'
• --drug-concentration: Two float values [MEKi, EGF]
• --output: Path for HDF5 results file
• --plot-output: Path for output plots
• --n-cells-plot: Number of cells to plot (population only)
• --skip-simulation: Skip simulation if results exist

Time Points

The simulation uses non-uniform time points (see src/main.py, startLine: 111, endLine: 119):

• Early phase: 0-600s, 24 points
• Middle phase: 600-3600s, 10 points
• Late phase: 3600-7200s, 5 points
• Pre-equilibration: -600-0s, 4 points

Output Files

Data Files

HDF5 files containing:

• Time series data
• Species trajectories
• Metadata (cell line, drug concentrations)

Plots

Two visualization options:

1. Basic Plotting (see src/plot_results.py, startLine: 10, endLine: 58):

• ERK Phosphorylation
• MEK Phosphorylation
• RAS Activity

2. Advanced Visualization (see src/visualization.py, startLine: 22, endLine: 52):

• Individual cell trajectories
• Population means
• Standard deviations
• Custom observable plotting

Model Parameters

Cell line-specific parameters (see src/models/old/parameters.py, startLine: 1, endLine: 23):

• Mutant cells: Higher baseline differentiation, stronger ERK dependence
• Wildtype cells: More epigenetic plasticity, slower differentiation

Dependencies

All required packages are installed via setup_pysb_env.sh:

• Core: numpy, scipy, pandas
• Visualization: matplotlib, seaborn
• Computation: cython, numba
• GPU Support: cudatoolkit, cupy
• Development: pytest, black, flake8

Citation

If you use this code in your research, please cite: [Citation information to be added]

SLURM Submission

Submit jobs using the provided SLURM script:

sbatch run_simulation.slurm

SLURM configuration:

• Partition: standard
• Memory: 16GB
• CPUs: 8
• Time limit: 12 hours

Visualization

Additional plotting tools are available (see src/visualization.py, startLine: 1, endLine: 53):

python src/plot_results.py \
    --results-file results/simulation_results.h5 \
    --n-cells 5

Post-Processing Analysis

For detailed analysis of simulation results, use the post-processing script:

python src/postprocess.py --input results/simulation_results.h5

This will:

• Generate detailed plots of species trajectories
• Save plots to results/analysis/species_trajectories.png
• Print statistics about the simulation data
• Show time evolution of first 8 species
• Display min/max concentrations for each species

The script provides:

• Time series visualization
• Basic statistical analysis
• Concentration ranges for each species
• Metadata from simulation conditions

Output location:

• Plots: results/analysis/species_trajectories.png
• Logging: Terminal output with statistics and file paths

Troubleshooting

1. Cython compilation errors:

conda activate pysb_env
conda install cython

2. Missing dependencies:

conda install -c conda-forge <package_name>

3. CUDA Limitations: Our model contains complex expressions and rules (see src/main.py, startLine: 74, endLine: 86) that are not compatible with CUDA-based solvers. Therefore, we use the CPU-based ODE solver (ScipyOdeSimulator) with Cython compilation for optimal performance. If you encounter CUDA-related errors:

• Use simulator_type='ode' instead of 'sde'
• Ensure Cython is properly installed
• The simulation will automatically use CPU-based computation

Note: While CUDA-based solvers can offer significant speedup for certain types of models, they are limited to simpler reaction networks without complex expressions. Our model's sophistication requires the more flexible CPU-based solver.

3. Initialize and push to GitHub:

# Initialize git repository
git init

# Add files
git add .gitignore README.md setup_pysb_env.sh src/ run_pysb.slurm

# Initial commit
git commit -m "Initial commit: PYSB simulation framework"

# Create new repository on GitHub through web interface
# Then link and push:
git remote add origin https://github.com/YOUR_USERNAME/pysb-simulation.git
git branch -M main
git push -u origin main

The code structure we're pushing is based on:

• Main simulation code (startLine: 1, endLine: 181 in src/main.py)
• Parameter handling (startLine: 1, endLine: 23 in src/models/old/parameters.py)
• Plotting utilities (startLine: 1, endLine: 58 in src/plot_results.py)

Would you like me to help with creating the GitHub repository through the GitHub API, or would you prefer to create it manually through the web interface?