Dr. Lai's research is directed toward understanding the fundamental mechanisms of gene regulation. We care about how protein complexes target the genome, how their binding changes across cell states, and how mis-regulation of these interactions produces disease phenotypes. DNA sequence, local and distal chromatin, and RNA all act cooperatively and antagonistically in determining when and how often a gene is transcribed. Resolving how these networks make decisions is a central bottleneck in connecting molecular perturbation to disease phenotype.

The Lai laboratory pairs wet-bench biochemistry with computational method development, combining high-resolution genomic assays with custom bioinformatic software. Contributions from the group and its collaborators include a single-base-resolution protein architecture of the budding yeast genome, a ChIP-exo screen of 887 transcription factor antibodies in human cells, and genome-wide characterization of RNA polymerase II initiation complex assembly. Projects are deliberately large in scale, generating thousands of unique (epi)genomic datasets across human, mouse, and yeast systems. The volume and dimensionality of these data require machine learning and explainable AI approaches deployed on national high-performance computing infrastructure.

A parallel emphasis of the laboratory is rigor, reproducibility, and access. The group develops and maintains open-source genomics software used by the wider research community including ScriptManager, PEGR, GenoPipe, and STENCIL. We also build science gateways that lower the technical barrier to advanced computation. Dr. Lai serves on the External Advisory Board for NSF ACCESS, the national research cyberinfrastructure program, and has directed shared genomics research infrastructure, working directly with physician scientists to apply epigenomic technologies and algorithmic approaches to biomedical specimens. Students are trained to work across biochemistry, molecular biology, and bioinformatics, and across a wide spectrum of heterogeneous and high-performance compute systems.