Cell-Cell-Communication

Summary

We provide code associated with the development of hybrid multiscale models of cell-cell communication. The models are developed in Julia for a specific gene regulatory network in hematopoiesis that controls myeloid cell fates; however the framework is general and can be applied in various contexts to gene regulatory networks linked to single cell phenotypes. The hybrid model consists of coupled continuous and discrete dynamical processes:

• Continuous gene regulatory dynamics inside cells (GATA1-PU.1 network), modeled with nonlinear differential equations as described previously (Chickarmane et al., 2009).
• Discrete cell-cell communication between cells, where signals sent between cells are modeled by a Poisson process.

In general, the form of the external signals can be defined by the user. In the paper, we consider two signals: consensus signals ("be like me") that recruit neighboring cells to commit to the same lineage, or dissensus signals ("be unlike me") that push neighbors to commit to the alternative lineage. Cell-cell signaling network topologies are defined by a matrix representation, where: 1 corresponds to a consensus signal; -1 corresponds to a dissensus signal; 0 corresponds to no signaling between cells.

To run the model

• Topology_Trajectories.jl - runs simulation of individual model trajectories
• Topology_Probabilities.jl - runs sets of simulations over a range of parameters (no noise added)
• Simple_Topologies_w_Noise_Probabilities.jl runs sets of simulations over a range of parameters with noise (either intrinsic or extrinsic) added. For large runs, we recommend using hpc/cloud resources
• Plotting_Trajectories_and_Probabilities.ipynb gives the code used to produce the plots in the paper, both for single trajectory simulations and approximate probability distributions for large sets of parameters.

Paper

A single-cell resolved cell-cell communication model explains lineage commitment in hematopoiesis

Megan K Franke and Adam L MacLean

bioRxiv: https://www.biorxiv.org/content/10.1101/2021.03.31.437948v1

doi: https://doi.org/10.1101/2021.03.31.437948

Dependencies

• DifferentialEquations - for differential equation modeling
• Distributions - for probability distributions
• DelimitedFiles - for I/O
• Plots, LsqFit for plotting and curve fitting

References

• V. Chickarmane, T. Enver, C. Peterson, Computational modeling of the hematopoietic erythroid- myeloid switch reveals insights into cooperativity, priming, and irreversibility, PLoS Comput Biol 5 (1) (2009) e1000268.
• J. Bezanson, A. Edelman, S. Karpinski, V. B. Shah, Julia: A fresh approach to numerical computing, SIAM Review 59 (1) (2017) 65–98.
• C. Rackauckas, Q. Nie, Differentialequations.jl–a performant and feature-rich ecosystem for solving differential equations in julia, Journal of Open Research Software 5 (1) (2017).