Beth K. Martin

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

The paper develops STEAM, an evolution-augmented deep learning model, to predict mammalian enhancer landscapes across species using comparative genomics and single-cell data.

Multi-scale dissection, compaction and derivatization of mammalian developmental enhancers

The paper systematically perturbs enhancer sequences to study their functional properties and robustness, revealing insights into regulatory sequence space and enhancer activity.

Retracing and rewriting the evolutionary trajectories of mammalian developmental enhancers

The paper studies the evolutionary changes of mammalian enhancers by analyzing orthologous sequences across multiple species and using deep learning to model functional changes.