augustinas1/CellCycle-RNAseq: Publication code: Code for paper: A. Sukys and R. Grima, "Cell-cycle dependence of bursty gene expression: insights from fitting mechanistic models to single-cell RNA-seq data" (2025).

The code is used to perform mechanistic model-based inference on scRNA-seq data for a population of mouse embryonic stem cells (mESCs), where the cell-cycle phase (G1, S or G2/M) and cell age θ (position along the cell cycle) are known for each cell. The processed dataset used throughout is uploaded on Zenodo, and is based on the original work by Riba et al. [1].

Structure
src - main Julia code used to build quantitative models of gene expression, perform maximum likelihood estimation and model selection, and compute the confidence intervals using profile likelihood.
analysis - scripts for the mESC dataset-specific analysis, considering cell age-dependent and cell age-independent mechanistic models using cell-age-resolved mRNA count data.
notebooks - Jupyter notebooks (written in Julia) used to explore the results and to generate the figures in the paper.
Mathematica - Mathematica notebooks used to solve the piecewise ODEs for the means and variances of the age-dependent models.
data - directory to save the generated files, such as model-specific fits.