National Institute of General Medical Sciences 1R35GM166209-01

Theory and infrastructure to expand the universe of statistical genetic models

The award funds a project at Carnegie Mellon University to develop advanced statistical genetic models and computational infrastructure to improve the accuracy and interpretability of genetic studies, enabling better understanding of complex traits and disease mechanisms.

Description

This project aims to develop statistical and computational tools that separate biological signal from social and environmental structure in large-scale human genetic studies. The work includes developing multivariate, inter-generational theory, corrected estimators for genetic parameters, diagnostics for covariance induced by assortative mating and vertical transmission, and inference procedures robust to outcome rescaling. Infrastructure development includes a public benchmarking suite, probabilistic-programming templates for biobank-scale inference, and graph-based linear-operator toolkits for efficient genetic analyses. The project emphasizes estimators with finite-sample guarantees, uncertainty quantification, and stress-testing against bias sources. All software and materials will be open and well-documented to lower barriers for non-specialists, aiming to bridge scalable statistical frameworks and realistic models of complex traits to maximize biological insights.

View original record