Understanding the functions of genes requires the investigation of the structure of their regulatory networks of interactions. Single-cell RNA sequencing (scRNA-seq) brings new challenges and opportunities to the study of such networks. Researchers at Texas A&M University have developed a machine learning tool for constructing and comparing single-cell gene regulatory networks. scTenifoldNet is a machine learning workflow built upon principal-component regression, low-rank tensor approximation, and manifold alignment for constructing and comparing single-cell gene regulatory networks (scGRNs) using data from single-cell RNA sequencing. Their algorithm complements and enhances the commonly used differential expression analysis by revealing differences between samples in the regulatory relationships among genes, rather than the expression level. The researchers anticipate that, by deciphering the complexity of data that surpasses human interpretative ability, scTenifoldNet can help achieve breakthroughs in understanding regulatory mechanisms underlying cell behaviors.
Overview of the scTenifoldNet Workflow