Hakhamanesh Mostafavi is a human geneticist whose research focuses on the genetic basis and biology of human complex traits and diseases. His lab combines modeling and analysis of large-scale genetic and genomic data to investigate how genetic differences give rise to phenotypic variation and how evolutionary forces shape heritable variation.
His work has developed conceptual frameworks for interpreting genome-wide association studies (GWAS), explaining why GWAS findings often show limited overlap with genetic effects measured in functional assays and why common-variant and rare coding variant analyses systematically prioritize different genes. His research has also examined the limits of polygenic prediction.
He received his Ph.D. in Biological Sciences from Columbia University, working with Molly Przeworski, followed by postdoctoral training with Jonathan Pritchard at Stanford University. He is currently an Assistant Professor in the Center for Human Genetics and Genomics and the Department of Population Health at NYU Grossman School of Medicine.
What genes do genetic association studies discover?
Most human traits are highly polygenic. Although genome-wide association studies (GWAS) have identified thousands of associated variants, most lie in noncoding regions, making it difficult to identify the underlying genes.
Two main strategies are used to map genes from association signals. One integrates GWAS with eQTLs (genetic effects on gene expression) to link noncoding variants to regulatory effects. The other tests the burden of rare, disruptive protein-coding variants to implicate genes directly.
Our work shows that these approaches sample different slices of a shared pool of causal genes. GWAS broadly capture this pool, but gene ranking is largely random due to genetic drift. GWAS-eQTL integration preferentially implicates genes that are less constrained by selection and have simpler regulatory architectures, while rare variant burden tests prioritize genes whose disruption has more trait-specific and less pleiotropic effects.
These differences reflect the distinct biases of each mapping strategy, each revealing a different but complementary aspect of trait biology. As such, all remain central to future efforts to map the genetic basis of complex traits.