Decoding transcriptional regulation via a human gene expression predictor

Transcription factors (TF) regulate cellular activities via controlling gene expression, but a predictive model describing how TFs quantitatively modulate human transcriptomes was lacking. We constructed a universal human gene expression predictor and utilized it to decode transcriptional regulation. Using 1613 TFs′ expression, the predictor reconstituted highly accurate transcriptomes for samples derived from a wide range of tissues and conditions. The predictor′s broad applicability indicated it had recapitulated the quantitative relationships between TFs and target genes ubiquitous across tissues. Significant interacting TF-target gene pairs were then extracted from the predictor and enabled downstream inference of TF regulators for diverse pathways involved in development, immunity, metabolism, and stress response. Thus, we present a novel approach to study human transcriptional regulation following the ″understanding by modeling″ principle.