Zhonghua Liu

This study applies the RNAGAN foundation model to nasopharyngeal carcinoma to identify metastasis markers and mechanisms, revealing a protective role of adaptive immune responses over innate inflammation.

RNAGAN is an AI tool that enhances transcriptomic analysis with interpretability and small data capability through a GAN structure trained on human RNA-seq data.

CoxMDS identifies causal mediators in DNA methylation studies with survival outcomes, maintaining FDR control and higher power than existing methods.

HILAMA is a method for high-dimensional mediation analysis with latent confounding, showing improved FDR control and power in simulations and applied to ADNI proteomics-radiomics data.

This study generates gene-modified cattle and sheep using haploid stem cells with a modified iCHI approach that enhances embryo development.

This study evaluates how imputation from large-scale sequencing data improves the power of rare variant association studies, finding that TOPMed imputation increases detected variants and genes compared to whole-genome sequencing.

This paper introduces inDecay, a few-shot learning system for predicting CRISPR editing outcomes in cell lines and zygotes, achieving high accuracy with minimal data.

The paper introduces trans-PCO, a method for trans-eQTL mapping that improves detection of genetic variants affecting gene regulatory networks, identifying 14,985 high-quality trans-eSNP-module pairs.

ImmuneMirror is a machine learning pipeline and web server for neoantigen prediction, showing high accuracy and identifying potential therapeutic targets in cancer.

The paper investigates the shared genetic etiology between chronic diseases and heart failure risk, finding that longer leukocyte telomere length mediates a small portion of the asthma PRS effect on HF risk.

The paper develops MR-SPI, a Mendelian randomization method integrated with AlphaFold3 to identify causal protein biomarkers and predict 3D structural changes in Alzheimer's disease.

The paper identifies shared genetic loci between COVID-19 outcomes and rheumatoid arthritis (RA) or systemic lupus erythematosus (SLE) using cross-trait analysis and Mendelian randomization, finding increased RA genetic risk correlates with higher COVID-19 hospitalization risk.

Grace-AKO improves variable selection in high-dimensional genomic data by integrating gene network structures, achieving better finite-sample FDR control and model stability.

CoxMKF identifies DNA methylation CpG sites mediating environmental exposure effects on survival outcomes while controlling FDR in high-dimensional settings.

This study identifies metabolic features in mouse pre-implantation embryos and shows that NAD+ synthesis is essential for development and regulates minor ZGA through H3K27ac erasure.

ERStruct estimates the number of top principal components from whole-genome sequencing data using parallel computing and GPU acceleration.

This study establishes a new delayed implantation model in mice using letrozole and estradiol to investigate the effects of implantation delay on late pregnancy events.

The paper develops a new statistical test, DACT, to detect mediation effects in genome-wide epigenetic studies and applies it to identify DNA methylation sites mediating smoking's effect on lung function.