Abstract
Genome-scale gene expression analysis has become a standard approach for discovering biomarkers and understanding molecular mechanisms. Recent advances in omics technologies now enable investigations beyond conventional case-control comparison and standard gene-centric differential expression (DE) analyses. In this review, we highlight conceptual and methodological advances in using transcriptomic and multimodal omic data to elucidate diverse mechanisms of gene expression. We first provide a comprehensive overview of different types of gene expression study designs, along with suitable statistical testing, as well as key considerations. We then describe strategies for inferring gene co-expression and regulatory networks, with particular emphasis on context-specific network models and machine learning methods that capture the multifactorial nature of gene expression regulation. Finally, we present perspectives on emerging modalities such as single-cell and spatial transcriptomics, which enable unprecedented resolution in mapping regulatory complexity. We envisage that the concepts and examples described here will raise awareness and encourage the application of advanced network-based analyses.