Jia Wen

The paper maps cis- and trans-eQTLs in blood samples to infer gene regulatory networks and prioritize trait-relevant pathways.

EPIC-unmix improves cell-type specific analysis from bulk RNA-seq by integrating single-cell reference profiles, identifying novel genes and eQTLs in Alzheimer's disease.

This study introduces VAE-PRS, a deep learning model that improves polygenic prediction for blood cell traits by capturing interaction effects and outperforming traditional methods in 14 out of 16 traits.

This study integrates genetic analysis with single nucleus RNA-seq to identify cell-type specific transcriptomic perturbations in schizophrenia, highlighting mitochondrial dysfunction and protein modification as key factors.

The paper performs GWAS on cardiovascular outcomes in CKD patients, identifying stroke-associated SNVs but not replicating CHD findings, and finding genetic correlations with published GWAS.

The paper identifies eQTL and sQTL in African American blood cells, finding many rare variants undetected in European ancestry studies and creating an open database for results.

The paper identifies functional CREs in iPSCs and neurons and links them to neuropsychiatric disease risk variants.

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

DeepGWAS enhances GWAS signals for neuropsychiatric disorders using deep neural networks, improving locus detection in schizophrenia and validating results in other disorders.

This study introduces S-BEAM, a semi-supervised ensemble method to prioritize non-coding variants and their target genes in microglia for Alzheimer's disease, identifying 11 risk variants and proposing 37 for further study.

The paper finds that rare copy number variants (CNVs) are more common in treatment-resistant psychosis patients compared to a general schizophrenia cohort, with duplications at 15q11.2-q13.1 and 16p11.2 linked to treatment resistance.

This paper compares eight ST deconvolution methods using four real datasets to evaluate their performance in inferring cell type composition from spatial transcriptomics data.

The paper sequences ancient genomes of Sussemionus, re-dating its extinction to the late Holocene and revealing genetic divergence and admixture events in the Equus lineage.

THUNDER is a deconvolution method that infers cell type proportions from bulk Hi-C data, outperforming existing methods in accuracy.

This study identifies 557 new gene-trait associations for hematological traits using transcriptome-wide association analysis in UK Biobank data.

The paper improves genotype imputation for CF patients using the TOPMed reference panel and a CF-specific imputation panel, enhancing marker coverage and PRS accuracy.

This study prioritizes functional elements of ~148,000 GWAS variants for blood cell traits using a web tool, showing prioritized variants are more likely to be experimentally validated.

The paper identifies novel genetic associations for biomarker traits in non-European UK Biobank populations, highlighting missed signals in European-centric analyses.

The paper identifies cell type-specific super-interactive promoters in blood lineage cells using Hi-C data and shows their association with ATAC-seq peaks and GWAS variants.

SnapHiC identifies high-resolution chromatin loops from scHi-C data and demonstrates cell-type-specific chromatin architecture in mouse and human cells.

This study characterizes 3D epigenomic landscapes during human corticogenesis, revealing chromatin looping and super interactive promoters involved in lineage-specific gene regulation.

This study identifies a large number of structural variants in human genomes using multiple sequencing technologies and presents them as a gold standard for future studies.