Yung-Han Chang

The paper shows that RNA-seq quantification methods and reference annotations significantly impact eQTL analysis, colocalization, and TWAS results across multiple tissues and phenotypes.

ICONIC is an R package that integrates instrumental variable and negative-control methods for causal inference in multi-omic studies, offering diagnostics and benchmarks for mediation analysis.

This study performs a cell-type-aware transcriptome-wide association study (TWAS) of mammographic density phenotypes, identifying novel genes and prioritizing known GWAS loci through cell-type-specific gene expression analysis.

This paper introduces a formal statistical replication analysis to reduce false positives in lung cancer GWAS, finding it significantly lowers false discovery rate and improves polygenic risk score efficiency.

This study identifies genetic risk mechanisms for six cancers by analyzing isoform-level expression data, finding more associations and greater heritability explained compared to gene-level analysis.