Projects

Clustering alignment to complement and advance single-cell clustering

I extend the clustering alignment framework to the single-cell analysis pipeline. Specifically, I develop new metrics for feature-level interpretation to identify cluster-informative genes, and provide measures of clustering consistency for comparing results from different clustering models. The alignment-enabled single-cell analysis pipeline streamlines robust clustering results evaluation, straightforward model comparison, and gene selection for downstream analyses.

ML-based methods for multivariate genetic association analysis

I devise methods to analyze genetic association signals and better characterize the genetic architecture of complex traits. I developed ML-MAGES, which uses neural-network-based supervised learning to account for LD-induced inflation in single-trait, SNP-level genetic associations, uses an infinite mixture model to categorize multi-trait association patterns, and aggregates and visualize association signals at gene-level. This approach substantially improves computational efficiency with large LD matrices and nonlinear inflation, achieves effect-size shrinkage comparable to or better than existing methods, and enables data-driven inference of an arbitrary number of association patterns in high-dimensional multi-trait settings.

Alignment of inferred latent ancestries in multiple population structure analysis results

I tackle cluster misalignment in population structure inference, where the inferred latent ancestries (clusters) vary across runs and choices of number of clusters (K>). I developed Clumppling to align multiple results, consolidate distinct solutions (modes), and clarify correspondence among inferred clusters. My tool visualizes the inferred population structure across K with multipartite graphs of bar plots and explicit alignment connections, enabling consistent, interpretable comparisons.

Mathematical properties of allele-sharing dissimilarities

I derive and analyze the mathematical properties of variants of allele-sharing dissimilarity measures (both within and between populations) as functions of allele-frequency distributions, providing guidance on their appropriate use in empirical studies.

Adaptation of genetic evolution methods to studying cultural evolution

Cultural evolution is the study of how socially learned behaviors, ideas, technologies, and norms change over time by processes analogous to biological evolution, but operating through social learning rather than genes. I study cultural variation using methods inspired by biological evolution, including the extraction of hierarchical structure in cultural variation and the intergenerational transmission of cultural traits.