Date: Thursday, November 13, 2025
Time: 11:45 a.m. to 12:45 p.m.
Location: G01 Gates Hall
Title: How Life Learns
Abstract: Genomics is changing fast. For years we worked with short reads and fragments, and we could only guess at how genomes really evolve. Now, with complete assemblies and pangenome models, we can see the whole picture—across species, across people, and across every part of the genome. Getting there has meant building new computational systems: traditional genomics tools compared everything to a single reference, but pangenomics requires tracking variation and many-way relationships across hundreds of complete genomes simultaneously, fundamentally changing the scale and structure of the problem.
I'll share a few glimpses of what that looks like: a human pangenome reference built from 466 genomes, unexpected recombination between the short arms of human acrocentrics, structural changes that track biological adaptation to shifts in human culture, hidden links between subtelomeric regions across chromosomes, and a first-ever view of whole-genome evolution in 581 vertebrates. New algorithms on sequences and graphs, applied at the scale of thousands of compute nodes, drive our advance. Together, these views suggest the field has entered a new phase that promises a complete understanding of how life copies, rearranges, and learns.
Bio: Erik Garrison is an Associate Professor in the Department of Genetics, Genomics and Informatics at the University of Tennessee Health Science Center, where he leads a research group developing computational methods for pangenomics and genome evolution. He is a core member of the Human Pangenome Reference Consortium (HPRC) and led the development of vg, the variation graph toolkit that has become foundational infrastructure for graph-based genomics. His open-source contributions include the PanGenome Graph Builder (pggb) pipeline and related tools for constructing and analyzing pangenome graphs without reference bias. His work has contributed to landmark publications including the first draft human pangenome reference, the discovery of recombination between heterologous human acrocentric chromosomes, the first complete human genome assembly (T2T-CHM13), and phase one of the Vertebrate Genomes Project. Prior to his appointment at UTHSC, Garrison was a postdoctoral fellow at the UC Santa Cruz Genomics Institute. He received his Ph.D. in Genomics from the University of Cambridge in 2019, where he was a student at the Wellcome Sanger Institute advised by Richard Durbin. His doctoral thesis, "Graphical pangenomics," established foundational methods for using sequence variation graphs in genome inference and alignment. He received a B.A. in Social Studies from Harvard University in 2006, where he was an Undergraduate Fellow at the Harvard Institute for Quantitative Social Science.