Evolutionary transfer learning enables organism-wide inference of mammalian enhancer landscapes

Checked 1 Oct 2026. Open license (CC-BY, CC-BY-SA, CC0, public domain): full text indexed and used in synthesis.

Understanding and modeling how a single human genome encodes gene regulatory programs for thousands of cell types remains a central challenge in genomics and machine learning. However, most human cell types emerge during embryonic, fetal, and pediatric development which are inaccessible to comprehensive molecular profiling. To overcome this, we hypothesized that the mismatch in evolutionary rates between cis -acting enhancers (fast) and the trans -acting regulatory programs that interpret them (slow) creates an opportunity for ‘evolutionary transfer learning’. Specifically, models trained to predict cell type-specific enhancers in one species should generalize to the orthologous cell types and genomes of related species. To test this, we generated a single-cell atlas of chromatin accessibility spanning mouse embryonic day 10 (E10) to birth (P0). Using combinatorial indexing(Cusanovich, 2015), we profiled 3.9 million nuclei from 36 staged embryos, resolving genome-wide accessibility in 36 cell classes and 140 cell types. With the goal of identifying distal enhancers for all cell classes, we then trained a series of multi-output deep learning models (CREsted(Kempynck, 2025)), each addressing limitations of the preceding approach. An ‘evolution-naive’ model achieved strong performance on heldout peaks, but exhibited two failure modes during genome-wide inference: overprediction at tandem repeats and conflation of promoter and distal enhancer grammars. An ‘evolution-aware’ model resolved both by regrouping accessible regions based on functional coherence across syntenic orthologs, but failed to generalize across species — suggesting insufficient sequence diversity during training. Finally, STEAM (Synteny-aware Transfer learning for Enhancer Activity Modeling), our ‘evolution-augmented’ model, expanded the training corpus to include enhancer orthologs from up to 241 mammalian genomes (Zoonomia(Christmas, 2023)) in a synteny-supervised manner, increasing the effective data scale by 195-fold. Despite increased label noise, STEAM learns a regulatory grammar that generalizes across mammals, supporting inference of 32 cell class, genome-wide enhancer maps for the human and mouse genomes ( HumMus ), as well as 239 additional mammalian genomes(Christmas, 2023) ( BabaGanoush ). Together, our results unify advances in single-cell profiling, deep learning, and comparative genomics into a framework for the evolutionary transfer learning of noncoding regulatory grammars. More broadly, our work supports the view that model organisms and evolutionarily diverse genomes are indispensable resources for accelerating the AI-enabled exploration of human biology.

None yet.

• DDLS pack: evolution and biodiversity