Deciphering the regulatory landscape of untranslated regions across cell types

The National Institute of General Medical Sciences awarded Whitehead Institute for Biomedical Research $125,000 on September 1, 2026, under the Biomedical Research and Research Training program (CFDA 93.859) to investigate cell type-specific regulatory mechanisms in untranslated regions of messenger RNAs and develop machine learning models for predicting the functional impact of genetic variants across different cellular contexts.

The research probes untranslated region (UTR) activity across three physiologically relevant cell types—stem cells, neurons, and cardiomyocytes—to elucidate how identical sequences produce distinct regulatory effects in different cell types. The three project aims include investigation of mechanisms underlying the most biologically significant UTRs; systematic mapping of RNA-binding proteins associated with the top 100 UTRs to identify molecular determinants of UTR-mediated cell type specificity; and development of a machine learning model that predicts functional impact of genetic variants in UTRs across cell types with applications for forecasting UTR activity.

The award funds research at Whitehead Institute's Cambridge, Massachusetts location. The period of performance runs from September 1, 2026, through August 31, 2028.

Abstract

Understanding cell type-specific protein expression is fundamental to developmental biology and disease mechanisms. While untranslated regions (utrs) of messenger rnas can exert profound, cell type-specific effects on protein expression by regulating RNA stability, translation, and localization, our understanding of UTR regulatory regions and their physiological functions remains incomplete. Current UTR mapping efforts lack comprehensiveness and rely on screening approaches that fail to recapitulate the cell type-specific, physiological, and translation-limiting conditions under which utrs regulate expression, leaving the regulatory landscape largely unexplored. To address these limitations, i have developed a systematic UTR screening strategy that probes all human and viral utrs at single copy in three physiologically relevant cell types: stem cells, neurons, and cardiomyocytes. These screens have uncovered thousands of regulatory elements with substantial cell type- specific effects, many controlling genes that are functionally critical in their respective cellular contexts. Leveraging this comprehensive UTR activity dataset, i propose to investigate UTR cell type-specific regulatory mechanisms and develop a machine learning model for predicting UTR activity across genetic variants and cellular contexts. My approach includes three aims. First, i will investigate the mechanisms underlying the three most biologically significant utrs to elucidate how identical sequences produce distinct effects across cell types. Second, i will systematically map RNA-binding proteins associated with the top 100 utrs to provide broader insight into the molecular determinants of UTR-mediated cell type specificity. Finally, i will develop a machine learning model that predicts the functional impact of genetic variants in utrs across cell types, with applications for forecasting UTR activity in diverse cellular contexts and designing cell type-specific utrs for RNA therapeutic applications. Together, this work will achieve a comprehensive understanding of UTR regulation across cell types from both mechanistic and predictive perspectives. IT establishes a generalizable framework for systematic exploration of UTR regulatory networks through functional genomics-informed AI modeling, accelerating discovery across molecular biology and transforming our understanding of post-transcriptional regulation in development and disease.