Image

Decoding the Regulatory Genome with Language Models

Hosted by: Professor Leonie Quinn

Abstract

This talk presents recent advances from my lab in developing genomic and RNA foundation models to decode the regulatory grammar of the genome. I will introduce NucEL, the first ELECTRA-style genomic language model with single-nucleotide resolution, which achieves high interpretability, computational efficiency, and state-of-the-art performance on gene regulatory benchmarks. I will then describe RegLLM, an RNA foundation model that generalizes across multiple post-transcriptional tasks—including protein–RNA interactions, RNA modifications, stability, and translation efficiency—by integrating structural and evolutionary features. Complementing this, I will present CS-Fold, a novel deep learning framework that leverages phylogenetic modeling of compensatory mutations to improve RNA secondary structure prediction. Finally, I will highlight our ongoing efforts to develop a single-cell foundation model for predicting cell-type-specific expression dynamics and perturbation responses, moving toward a virtual cell framework. 

Biography

Associate Professor Jean Wen leads the Computational Biology of RNAs and Functional Genomics group at the John Curtin School of Medical Research (JCSMR), Australian National University (ANU). She earned her PhD from ANU, followed by postdoctoral training with Professor Anders Krogh at the Bioinformatics Centre, University of Copenhagen, and Professor Eric Lai at Memorial Sloan-Kettering Cancer Center (MSKCC) in New York. In 2018, she returned to ANU as an ARC Future Fellow to establish her independent research group. At the intersection of computational biology and RNA biology, her team develops advanced machine learning frameworks, including large language models (LLMs), to decipher the intricate regulatory language of RNA. Integrating these models with high-throughput technologies like single-cell and spatial transcriptomics, they explore diverse RNA-mediated mechanisms—such as mRNA isoform diversity, RNA-binding proteins, small RNAs, RNA modifications, and conserved structural RNAs—to understand how post-transcriptional processes govern gene expression, cellular identity, and disease pathogenesis. Associate Professor Wen currently serves as Deputy Director of the Shine-Dalgarno Centre for RNA Innovation and ANU Node Leader for the ARC Centre of Excellence in Mathematical Analysis of Cellular Systems (MACSYS).