Yang David Dai

PGD improves visualization of single-cell trajectories by smoothing cell-level features along pseudotime, enhancing continuity and structure in complex biological processes.

CHROME improves regulatory genomics predictions by integrating physically validated chromatin interactions into graph models, enhancing accuracy and interpretability for ChIP-seq, eQTL, and variant pathogenicity.

This study shows that adding fermentable fiber to a Western diet reduces indole levels and improves gut microbiota diversity and metabolic health in mice.

RAGulate improves the assessment of transcription factor-target relationships by integrating retrieval-augmented generation with a regulatory knowledge base to reduce hallucinations and enhance evidence faithfulness.

scRegulate infers transcription factor activity and gene regulatory networks from scRNA-seq data using a deep learning framework, showing improved accuracy and efficiency over existing methods.

MiMeNet is a neural network framework that improves prediction of metabolite abundances and identifies microbe-metabolite interactions, outperforming linear models in paired microbiome-metabolome datasets.

Meta-Signer identifies metagenomic signatures for disease association by aggregating feature rankings from multiple machine learning models.

PopPhy-CNN uses a phylogenetic tree embedded architecture to predict host phenotype from metagenomic data, outperforming traditional and baseline models.

This study uses conditional generative adversarial networks to simulate microbiome data, improving disease prediction through synthetic data augmentation in inflammatory bowel disease.

MeDEStrand improves estimating absolute DNA methylation levels from enrichment data by correcting CpG bias using a sigmoid function and processing strands separately.