911±¬ÁÏÍø

Computational Models for Genome Regulation and Bioinformatics Education

Kimmo Palin
University of Helsinki

Abstract: Gene transcription regulation is written in the genome with weak, distributed signals such as  conservation patterns across species, population-scale genetic variation and protein footprints on individual DNA molecules. Reading those signals requires principled computational models. This talk covers two decades of work on such models: comparative-genomic enhancer discovery that led to a colorectal cancer causing variant, a message-passing phasing algorithm for haplotype phasing of thousands of human genomes, and whole-genome somatic mutation and methylation analyses across multiple cancer types. The current focus is on long-read methods that stamp protein-DNA contacts onto continuous DNA molecules at single-nucleotide resolution. In the second part of the talk I describe my ideas for bioinformatics and health data science education: a curriculum focused on understanding the computational tools, with particular attention to their statistical foundations, reproducible data analysis, and the emerging challenges and possibilities posed by generative AI in scientific computing.

Bio: Kimmo Palin is a Researcher at the University of Helsinki, where he works on data-analysis of genome regulation, genetic variation, and cancer biology mostly in context of colorectal cancer and uterine leiomyoma. He holds a PhD in Computer Science (2007) and Title of Docent (2023) in Bioinformatics from the University of Helsinki. He had a postdoc at the Wellcome Trust Sanger Institute 2008-2012, where he developed a haplotype phasing algorithm for large founder populations and returned to Helsinki to work in the Academy of Finland Centre of Excellence in Tumour/Cancer Genetics Research 2012-2025. He has supervised five PhD and multiple MSc and BSc theses and lectured in the TRANSMED programme at the University of Helsinki. He has 55 publications, >9500 citations and H-index of 36.