Computing Minimal Boolean Models of Gene Regulatory Networks.

Area of Science:

• Computational Biology
• Systems Biology
• Genomics

Background:

• Gene regulatory networks (GRNs) model cellular dynamics and gene expression variability.
• Boolean networks are a simple model for GRNs, but often face challenges due to incomplete data and unknown rules.
• Experimental limitations and data noise complicate the accurate reconstruction of GRNs.

Purpose of the Study:

• To develop an algorithm for optimal GRN structure identification from experimental data.
• To identify noise-free network states corresponding to observed gene expression data.
• To address the computational complexity of GRN inference, especially for single-cell RNA-Sequencing (scRNA-Seq) data.

Main Methods:

• A novel methodology integrating experimental data for GRN structure search.
• Optimization of a linear objective function under linear constraints to find consistent network structures.
• A heuristic extension to efficiently process large scRNA-Seq datasets.

Main Results:

• Demonstrated effectiveness using simulated data.
• Validated the methodology on a publicly available scRNA-Seq dataset and its associated GRN.
• Successfully identified optimal GRN structures and noise-free states.

Conclusions:

• The developed methodology provides a robust approach for inferring GRNs from complex biological data.
• This work enhances the understanding of GRN dynamics and their biological functions.
• The approach is particularly beneficial for analyzing scRNA-Seq data, overcoming computational challenges.