A Mecom—Cdk6 toggle switch governs hematopoietic stem-to-multipotent fate transitions via distinct multistable landscapes

Citation

Please cite the paper

Martinez, J. A., Dey, A., McDermott, M., Rudolph, K. L., & MacLean, A. L. (2026). A Mecom–Cdk6 toggle switch governs hematopoietic stem-to-multipotent fate transitions via distinct multistable landscapes. bioRxiv. https://doi.org/10.64898/2026.06.26.734844

Overview

This repository contains

• ODE files to generate one- and two-parameter saddle-node bifurcations using AUTO funtion in XPPAUT.
• MATLAB codes for generating phase-plane digrams with fixed points, determinsitic trajectories and nullclines, for constructing most probable paths using minimum-action potential (MAP), simulating stochastic trajectories using SDE, constructing the stable, unstable manifolds
• Python codes for pre-processing single cell data and generating psuedocell over psuedotime plots of gene expression (for Mecom and Cdk6) and UCell Score

Requirements

• MATLAB (tested on version R2024a)
• Python (tested on version 3.11.15)
• All parameter values are included within respective code files unless otherwise specified

Figure 1

• environment.yml - Environment File for the Python codes
• Code_Fig1B_Pre_Processing.ipynb - Provides Pre-processing for Single Cell Data (yAL and yDR) as described in the methods
• settings_allRuns_clusterLabels.py - Helper Function to import clustering labels
• Code_Fig1B.ipynb - Generates Figure 1B and Fig S1
• Mecom_Cdk6_Model.ode - Generates the bifurcation in Fig 1C using AUTO in XPPAUT
• Code_Fig1D.m - Generates figure 1D
• Code_Fig1E.m - Generates figure 1E. Appropriately change the noise strength (sigma).
• Code_Fig1_SI_Nullclines_Trajectories.m - Generates Fig S2

Figure 2

• Mecom_Cdk6_Model.ode - Generates Fig 2B,C. Appropriately change the 'alpha_X' parameter to see the effect of pertubing Mecom self-activation
• Code_Fig2DE.m - Generates Fig 2D (nullclines) ,E
• Code_Fig2D.m - Generates Fig 2D (trajectories and fixed points)
• All of these are used to generate Fig S3

Figure 3

• Code_Fig3_SDE_StochData.m - Generates data for stochastic trajectories for SDE solved using Euler-Maruyama method. The stochastic trajectories are stored as a '.mat' file which is used to plot the cell states at the final simulation time (blue, yellow and red points) in Figure 3.
• Code_Fig3_FixedPoints_Manifolds.m - Generates fixed points and separatrices i Fig 3.
• Code_Fig3_MostProbablePaths.m - Generates most probable paths using Minumum-action path to plot Fig 3C, F.
• All of these are used to plot Fig S5

Figure 4

• Mecom_Cdk6_Model.ode - Generates the 2-parameter bifurcation in Fig 4A using AUTO function in XPPAUT
• cusp_detection.m - Generates the cusp-bifurcation points associated with Fig 4A
• Code_Fig4BC.m - Generates the phase-planes with fixed points, separatrices and MAPs for Fig 4B,C.
• All of these are used to generate Fig S6.

Figure 5

• Mecom_Cdk6_Model.ode - Generates the 2-parameter bifurcation in Fig 5B using AUTO in XPPAUT
• Code_Fig5_SDE_DdrugPerturb.m - Generates stochastic trajectories for SDE solved using Euler-Maruyama method. The output is stored as a '.mat' file to be used to plot the cell states at the final time in Fig 5A. Appropriately change 'ddrug' parameter to see the effect of virtual inhibition of CDK6 by the drug.
• Code_Fig5_FixedPoints_Manifolds.m - Generates fixed points, separatrices for Fig 5B
• All of these are also used to generate Fig S7