Trevor J. Hastie

The paper introduces StaBLE, a digital balance assessment tool validated against clinical scales, showing strong correlations and improved detection of fall risk.

CuGen is a GPU-accelerated framework for large-scale genomics that enables efficient GWAS and fine-mapping with robust confounding control.

This paper uses machine learning to classify leiomyosarcoma based on methylation signatures, finding differences in PD-1 pathway activation and immune cell composition between tumor subtypes.

The paper identifies tissue-specific transcriptional regulatory programs in rats undergoing endurance exercise training by integrating chromatin accessibility, DNA methylation, and gene expression data.

This paper introduces Combine, a framework for polygenic prediction in admixed populations using local ancestry and group lasso, showing improved performance over existing methods.

The paper develops a method to identify genetic predictors of cardiometabolic disease progression using time-to-event data from the UK Biobank, finding ten polygenic hazard scores that significantly predict disease progression.

The paper improves a method for inferring 3D chromatin structure from Hi-C data by introducing a new framework that addresses initialization issues and extends distribution models for contact counts.

This paper introduces a method to identify causal variants from genome-wide summary statistics, outperforming existing methods and discovering additional loci for complex traits.

This study profiles multi-omic changes in multiple tissues of rats during endurance exercise training, identifying molecular adaptations and sex-specific responses across time.

PathGPS uses GWAS summary data to identify genetic architecture by separating genetic components from environmental factors and finding gene-trait clusters associated with biological pathways.

The paper develops sparse polygenic risk scores for 813 traits using UK Biobank data, showing significant predictive performance and a correlation between variant count and performance.

The paper develops efficient solvers for large-scale regularized regressions on genetic data, reducing memory usage and improving computational speed.

This paper introduces a latent factor model for polygenic risk scores that decomposes genetic risk into interpretable components, showing improved interpretability over standard PRS for traits like BMI and myocardial infarction.

This paper introduces a scalable framework for sparse regression on large-scale data, demonstrating improved prediction performance on UK Biobank traits using the BASIL method and snpnet package.

This paper introduces a principal curve method for inferring 3D chromatin architecture from Hi-C data, showing its utility and reproducibility in reconstructing chromosome structures.

The paper introduces a group-regularized Cox regression method for survival analysis of rare events using common responses, showing improved prediction performance on UK Biobank data.

The paper develops a scalable sparse reduced-rank regression method for high-dimensional multitask learning, applied to UK Biobank data to generate polygenic risk scores.

The paper introduces a machine learning method that integrates immunological knowledge to improve predictive power from mass cytometry data.

The paper uses DeGAs to identify genetic components across 2,138 phenotypes in the UK Biobank, revealing novel adipocyte biology through variant and gene contributions to obesity-related traits.

UK Biobank genetics of 38 blood and urine lab tests yields 1,857 independent loci; combining biomarker and disease polygenic scores improves FinnGen prediction of incident renal failure and alcoholic cirrhosis.

This study evaluates seven wrist-worn devices for heart rate and energy expenditure accuracy in diverse conditions, finding significant variability and proposing validation standards.