Canceled: Anshul Kundaje: Decoding regulatory DNA with deep learning models

Speaker:
Anshul Kundaje (Stanford University, USA) 

Title: 
Decoding regulatory DNA with deep learning models

Abstract:
Transcription factors interpret complex cis-regulatory logic encoded in DNA to modulate gene expression. Deciphering this “cis-regulatory code” is essential for understanding how genes are controlled across cell types, individuals, and species, and for explaining how noncoding variation shapes traits and disease. Large-scale molecular profiling efforts now provide rich, cell-type–specific regulatory maps, offering an unprecedented opportunity to apply machine learning to decode sequence-function relationships. In this talk, I will present lightweight deep learning models coupled with interpretation frameworks that (1) uncover causal sequence syntax and its pleiotropy across molecular and cellular contexts, (2) detect and correct experimental biases, (3) infer fundamental biophysical parameters, and (4) prioritize functional regulatory variants underlying common and rare diseases. Our models achieve or surpass the performance of massive multi-task and self-supervised DNA language models across benchmarks including molecular QTLs from diverse ancestries, reporter assays, CRISPR perturbations, and fine-mapped GWAS loci. By systematically interpreting ~5,000 models trained on bulk and single-cell datasets spanning diverse fetal and adult contexts, we expose the remarkable complexity and context-specificity of regulatory lexicons, syntax and variation encoded in the human genome.

Lecture Series:
SysBio Lecture Series: AI for Systems Medicine