This repository allows to reproduce the figures of the paper Ventre et al. [1], which describes a method of trajectory inference for scRNA-seq data with lineage tracing. Note that some of the functions in these files, in particular the ones to simulate branching SDEs and to solve a convex minimization problem using Mean-field Langevin dynamics, are directly taken or adapted from two previously published papers, [2] and [3] respectively. This is of course always specified at the head of the files.
[1] Ventre, E., Forrow, A., Gadhiwala, N., Chakraborty, P., Angel, O., & Schiebinger, G. (2023). Trajectory inference for a branching SDE model of cell differentiation. arXiv preprint arXiv:2307.07687.
[2] Forrow, A., & Schiebinger, G. (2021). LineageOT is a unified framework for lineage tracing and trajectory inference. Nature communications, 12(1), 4940.
[3] Chizat, L., Zhang, S., Heitz, M., & Schiebinger, G. (2022). Trajectory inference via mean-field Langevin in path space. arXiv preprint arXiv:2205.07146.
First, clone this repository with
git clone https://github.com/eliasventre/LTreweightingDependencies are listed in pip.requirements. They can be installed with
pip install -r pip_requirements.txtThe times expected for running the python files on a personal laptop are the following:
Figure2.py : about 15 minutes,
Figure3.py: about an hour,
Figure4.py: about a minute,
Figure5.py: about 20 minutes,
Figure6.py: about two hours.
You should then be able to run the script for the figure you want to reproduce.