Stochastic Modeling of Biophysical Responses to Perturbation

Notebooks for reproducing all figures and analysis of single-cell datasets in the paper. All saved/processed data used for analysis can be found on CaltechData. Figure created with BioRender.com.

Getting Started

------ Tutorial for Inferring Condition-Specific Kinetics -----

See the condition_kinetics_example.ipynb

------ Tutorial for Predicting Kinetics in Combined Conditions -----

See the combinatorial_prediction_example.ipynb

------ Tutorial for Clustering Perturbed Populations -----

See the perturbation_populations_example.ipynb

------ Tutorial in generating U and S counts -----

See the get_data_example_notebook.ipynb.

For notebooks that run on Google Colab, you will see the Colab link included at the top of the notebook. Just click on the symbol.

An introduction to using Google Colab can be found here. Briefly, run each code cell by selecting the cell and executing Command/Ctrl+Enter.

#To install Monod (and meK-Means)

pip install monod

import monod
from monod import mminference #Function implementing meK-Means algorithm

Related Links

meK-Means tutorial for biophysical clustering of multimodal data.

All analyses utilize the Monod package for single-cell, CME-based inference.

Repo Directory Contents

1. notebooks

   • Jupyter notebooks for all analyses, with relevant main Fig or suppFig denoted.

2. scripts

   • Preprocessing scripts to convert FASTQs to loom files with U and S counts
   • For each dataset, processing order is fetch.sh --> count.sh --> concatLooms.py

3. metadata

   • Metadata files for datasets analyzed.

4. reference

   • Gene length transcriptome references for Monod

5. results

   • Saved results from parameter prediction, HOMER analysis, and genome location analysis.