Doug Speed

This paper introduces GenSep, a framework to quantify genetic separability of disease subtypes, showing current PRSs fall short of theoretical limits.

This study identifies genetic variants associated with healthcare costs across multiple systems, finding polygenic influences and large effects from specific loss-of-function variants.

SNPic uses topic modeling to cluster complex phenotypes by interpreting GWAS data as a corpus, identifying interpretable genetic modules and biological mechanisms.

StratGWAS improves GWAS power by incorporating phenotype heterogeneity while maintaining specificity, identifying more significant loci in depression and other traits.

The paper identifies genetic variants associated with spontaneous coronary artery dissection using multi-trait GWAS and gene-based analysis, revealing new loci and biological pathways.

LDAK-KVIK is a mixed-model association analysis tool that outperforms existing methods in power and efficiency for quantitative and binary phenotypes.

LDAK-PBAT detects gene pathways associated with complex traits using GWAS summary statistics, outperforming existing methods in sensitivity and specificity.

This study identifies 80 genetic loci and 61 effector genes associated with dilated cardiomyopathy through GWAS and integrates transcriptomic data to uncover biological pathways and cellular states involved in the disease.

This paper evaluates the impact of participation bias in the UK Biobank by applying inverse probability weighting to genome-wide association analyses, revealing significant distortions in genetic correlations and Mendelian Randomization estimates.

LDAK-GBAT is a gene-based association testing tool that identifies more genome-wide significant genes than existing methods using summary statistics from large-scale studies.

This paper presents a publicly available privacy-preserving benchmark for polygenic prediction using summary-level data, enabling accurate model performance estimation without access to individual-level genetic data.

This paper applies genome-wide association, prediction, and heritability analyses to bacterial data, identifying genetic factors influencing antibiotic resistance and within-host survival in Streptococcus pneumoniae.

The paper identifies TSPAN8 variants associated with metabolomic regulation and metabolic diseases using CNV analysis and functional characterization in human induced pluripotent stem cells.

This paper develops and compares genetic prediction tools that incorporate heritability models, showing improved accuracy for complex traits using individual-level and summary statistics data.

The paper estimates the proportion of phenotypic variation captured by DNA methylation markers in 400 traits and finds that most complex traits show little association with these markers, though some like smoking have moderate capture.

This study estimates narrow-sense heritability in admixed populations using simulations and proposes a PCA-based adjustment to improve accuracy.

The paper evaluates heritability models using summary statistics from genome-wide association studies and finds the Baseline LD Model is most realistic but can be improved by incorporating features from the LDAK Model.

This paper shows that LDSC and SumHer can misestimate heritability due to confounding bias and covariate adjustments in GWAS data.

This study uses Mendelian Randomization to investigate the causal relationship between body fat, non-fat mass, height, and depression, finding that fat mass and short stature increase depression risk while non-fat mass does not.

The paper introduces SumHer, a method for estimating SNP heritability from summary statistics, showing it provides more accurate results than LDSC and reveals lower SNP category enrichments in complex traits.

This study uses Mendelian randomization to investigate if neuroticism causally increases the risk of depression.

The paper challenges the validity of S-LDSC by showing it produces unreliable enrichment estimates and defends the LDAK Model as more accurate for SNP heritability analysis.

The paper investigates genetic correlations between psychiatric and immune-related phenotypes using GWAS summary data, finding significant shared genetic variants.

The paper re-evaluates SNP heritability estimates for 19 human traits using an improved model, finding higher heritability estimates compared to existing methods and lower contribution from DNaseI hypersensitivity sites.

The paper investigates how structural variations in yeast genomes affect gene expression and quantitative traits, finding that copy number variants significantly influence phenotypes and reproductive isolation.