Samyak
Samyak
Revere High School
Revere, MA
Mentor:
Maheak Vora
Mercader Lab
Samyak spent his summer using computational genetics to investigate why the same genetic variant can have different relationships with disease across populations. His project focused on type 2 diabetes (T2D) and the importance of ancestry when interpreting genome-wide association studies, or GWAS.
Most GWAS analyze variants using an additive model, in which each additional copy of a risk allele is assumed to incrementally change disease risk. Samyak and Christelle Moise investigated a different possibility: some variants may behave more strongly under a recessive model, in which an effect becomes apparent primarily when an individual inherits two copies of the relevant allele. They asked whether variants showing ancestry-related heterogeneity were enriched for these recessive effects.
Samyak worked with large GWAS summary-statistics datasets, matching variants by genomic coordinates and alleles and comparing ancestry-related heterogeneity statistics with results from an internal recessive GWAS. This required computational data analysis and statistical reasoning, including the use of transformed P-values, correlation analyses, allele frequencies, odds ratios, confidence intervals, and forest plots.
Across 1,244 shared variants, the team found a modest positive correlation between ancestry-related and recessive heterogeneity. They then prioritized PAX4, an established T2D locus enriched in East Asian populations. PAX4 showed its strongest recessive association in East Asian populations, and cohort-level analyses supported the consistency of that signal across multiple East Asian datasets.
Samyak learned how data harmonization, statistical models, allele frequencies, P-values, odds ratios, and forest plots work together to reveal genetic effects that may differ across ancestry groups.
His project demonstrates why diverse populations are essential to human genetics. By studying how genetic effects vary across ancestry groups and inheritance models, his work contributes to a more complete-and ultimately more equitable-understanding of the genetic architecture of type 2 diabetes.
Samyak captured the experience in characteristically vivid terms, describing BSSP as “no mere crucible of research, but a noble fountain of wisdom and a forge for bonds that shall outlast the fleeting years.” His reflection speaks to both sides of the summer: the intellectual challenge of computational genetics and the relationships built while doing that work.
