Here, we introduce Refate, a computational framework that integrates large-scale multimodal single-cell atlas data from humans and mice and a repertoire of six drug databases for identifying genes with high cell propensity and subsequently predicting chemical compounds that target GRNs for converting cells from a starting type to a target type with minimum input from users.

● Refate quantifies genes for cell-fate conversion using multimodal single-cell atlases

● Refate uncovers gene regulatory networks (GRNs) underlying cell-fate conversion

● Refate identifies chemical compounds that target GRNs for cell-fate conversion

● Validation of Refate identified chemical compounds for hESC to hCNCC conversion