Adrian L. Sanborn

ATOM-1: A Foundation Model for RNA Structure and Function Built on Chemical Mapping Data

ATOM-1 is an RNA foundation model trained on chemical mapping data, enabling state-of-the-art RNA prediction tasks through small probe neural networks.

Simple biochemical features underlie transcriptional activation domain diversity and dynamic, fuzzy binding to Mediator

The paper identifies 150 transcriptional activation domains in yeast, shows their binding to Mediator, and develops a neural network to predict ADs across species.

Cohesin Loss Eliminates All Loop Domains, Leading To Links Among Superenhancers And Downregulation Of Nearby Genes

This study shows that cohesin loss eliminates loop domains, causing superenhancer links and gene downregulation, which are reversible upon cohesin restoration.