Inference of Gene-Gene Interaction and Its Function

The complex functions of cells are achieved through the interactions of diverse molecules and genes, and the analysis of gene interactions is essential to understanding the circuit of biological systems. Computational biology, machine learning, and artificial intelligence are becoming increasingly crucial for extracting insight into complex gene interactions from large-scale and complex data. In this chapter, we introduce computational methods for inferring gene-gene interactions and analyzing the function of genes. First, we briefly review the approaches and characteristics of gene-gene interaction inference algorithms. Then, we present a step-by-step protocol for gene network construction and analysis using CellOracle. This method learns regulatory patterns from single-cell gene expression and epigenomic data to construct transcriptional gene-gene association network models. One of the unique features of this method is the simulation of cell differentiation, which we call "in silico gene perturbation." This approach allows simulation of the dynamics of cell identity upon gene perturbation, facilitating the utilization of gene network models and interpretation of gene function.