Fig. 1

Biomodelling.jl workflow: (1) defining the gene regulatory network topology, interactions and parameters related to gene expression (2) choosing the number of cells and parameters related to cell population such as cell volume control and division noise, (3) couple a stochastic simulation algorithm of biochemical reactions with cell growth, size and division and simulate the cell population, (4) track genes or cells of interest and finally (5) save and export the data in a matrix similar to scRNA-seq data format. (I) synthetic data are generated using Biomodelling.jl using different sparsities as described in the Methods sect. "Network topology, sparsity and simulation" , (II) the obtained data are downsampled, then imputation and network inference are performed as described in Methods sect. "Downsampling, scaling and imputation" and " Network inference algorithms" , finally (III) network inference algorithms predictions are compared with the GT network using metrics presented in Methods sect. "Network inference performance evaluation"