Steven Gazal

This paper introduces CT-FM and CT-FM-SNP methods to identify causal cell types and variants for diseases, finding most SNPs affect traits via a single cell type set.

This paper introduces an annotation-based gene-peak linkage method that improves gene expression prediction from chromatin accessibility in human kidney multi-omics data.

The paper presents an isoform-level lung cell atlas using single-cell long-read RNA-sequencing, revealing genetic effects on isoform regulation that improve cell-type classification and identify novel variants associated with lung cancer risk.

The paper evaluates genetic-ancestry inference methods from single-cell transcriptomic data, showing ADMIXTURE provides accurate results despite challenges in variant calling.

The paper introduces a machine-learning framework to model disease effect sizes from GWAS data, identifying constrained regulatory variants as key functional components and improving variant prioritization.

This paper introduces PHBC to estimate pleiotropic heritability across diseases, finding that about half of common disease heritability is shared with other diseases.

jaxQTL improves sc-eQTL mapping by using count-based models, identifying more eGenes and explaining more SNP-heritability than bulk-eQTLs.

This paper detects widespread directional selection in ancient DNA data, identifying 347 genome-wide significant loci and linking them to traits like body fat, skin color, and cognitive performance.

The paper introduces SuShiE, a model that improves multi-ancestry fine-mapping of cis-molQTLs, identifying more genes and prioritizing functionally enriched variants.

The paper identifies proteins and regulatory elements associated with vocal learning by analyzing bat neural activity and epigenomic data.

The paper introduces LDSPEC to estimate correlations between causal disease effects of proximal SNPs, finding significant correlations that vary with distance, allele frequency, LD, and functional annotations, implicating stabilizing selection.

The paper identifies ancestry-specific regulatory and disease genes by analyzing single-cell RNA-seq and GWAS data, showing cell-type-specific gene-by-environment interactions.

TOGA integrates gene annotation and orthology inference across hundreds of genomes, improving ortholog detection and creating large comparative gene resources.

A machine learning approach identifies enhancer-phenotype associations across mammals, revealing links to brain size and behavior evolution.

The paper curates TE content in 248 mammalian genomes, finding trends in recent TE accumulation and associations with dietary habits.

A 240-species placental mammal alignment resolves constraint to single bases, estimating at least 10.7% of the human genome is constrained, with 48.5% of constrained bases unannotated.

The paper uses evolutionary constraint scores to identify functional genomic regions and finds they are enriched for variants associated with common diseases.

The paper identifies bats as a hotspot for horizontal transfer of DNA transposons, finding over 200 putative elements and shared transposons across mammals and reptiles.

FactorGo identifies pleiotropic components across thousands of human diseases and traits using GWAS summary statistics, outperforming existing methods in capturing latent factors.

This study assesses the contribution of Neanderthal introgression to human complex traits using UK Biobank data, finding significant but depleted effects compared to modern human variants.

The paper links SNPs to genes using functionally informed strategies, revealing enhancer-related and master-regulator genes contribute to autoimmune disease heritability.

The paper develops a combined SNP-to-gene linking strategy (cS2G) to improve the identification of disease-causing genes and assess disease omnigenicity using GWAS data.

This paper introduces PolyPred and PolyPred+ to improve cross-population polygenic risk score accuracy by addressing LD, MAF, and causal effect size differences using fine-mapping and non-European training data.

The paper partitions gene-level heritability by allele frequency to identify disease-relevant genes with heritability from rare variants, finding 370 gene-trait pairs not detected by existing methods.

The paper estimates the fraction of disease heritability mediated by genes linked to SNPs without using eQTLs, finding that the closest gene mediates 27% of heritability and some genes outside the top ten also contribute significantly.

The paper identifies environmental modulators of genetic risk factors for complex traits by integrating perturbation data with GWAS.

The paper integrates single-cell RNA-seq, epigenomic maps, and GWAS data to identify disease-critical cell types and processes across multiple diseases and tissues.

This study creates a single-cell and spatial atlas of autopsy tissues from COVID-19 patients, revealing altered cellular responses and potential therapeutic targets.

The paper develops a method to identify population-specific causal effect sizes in functionally important regions using trans-ethnic genetic correlations.

This paper introduces cov-LDSC, a method to estimate heritability and its enrichment in tissue-specific gene sets in admixed populations, showing high concordance across populations but population-specific differences for some traits.