Yumei Li

The paper characterizes nucleosome organization across 23 cancer types using ATAC-seq and shows its utility for cancer classification and detection using cell-free DNA.

cfMIND improves non-invasive disease detection by analyzing read-level methylation signals from cell-free DNA, achieving high accuracy even at low sequencing depth and across diverse cancers.

CellBin processes spatial omics data to create cell-level matrices, outperforming existing methods across multiple platforms and omics types.

This study develops a mouse model of rhodopsin sector RP and shows that dim red light and S1PR agonist treatment can prevent light-induced photoreceptor degeneration.

This study performs single-cell RNA sequencing on primary human trabecular meshwork cells in vitro to compare gene expression with in situ tissue, providing a public resource for studying cellular identity and translatability.

cfDecon improves cfDNA deconvolution accuracy and interpretability using deep learning, outperforming existing methods in simulation and clinical datasets.

The paper introduces a sample multiplexing method for scRNA-seq to improve resolution of sample origin and phenotypic data in mouse retinal ganglion cells.

This study benchmarks five methylation-based methods for cell type deconvolution of plasma cell-free DNA using deep whole-genome bisulfite sequencing data from 35 human cell types.

The paper shows SOX7 deficiency causes ventricular septal defects by impairing endocardial-to-mesenchymal transition through reduced Wnt4 and Bmp2 expression in developing atrioventricular canals.

The paper identifies CoRSIVs with strong genetic control of methylation and links them to transposable elements and disease.

The paper identifies cell type-specific genetic effects on gene regulation in human retina using single-cell multiomics data and finds that trans-factors, not cis-elements, drive these effects.

The paper re-evaluates differential expression methods on semi-synthetic RNA-seq data without normalization, finding the Wilcoxon rank-sum test still outperforms dearseq in FDR control and power.

The paper shows that popular differential expression methods like DESeq2 and edgeR have high false discovery rates in population-level RNA-seq data, recommending the Wilcoxon rank-sum test instead.

The paper benchmarks PCA against hidden variable inference methods for QTL mapping, finding PCA to be faster, better-performing, and more interpretable.

The paper identifies an atlas of 3QTLs associated with alternative polyadenylation across 46 tissues, showing they explain 16.1% of trait-associated variants and contribute to disease independently of gene expression.