Andrew Young Koh

This paper introduces a supervised contrastive learning framework for multi-omics prediction models, showing improved accuracy over single-omics approaches in type 2 diabetes and high-fat diet studies.

This paper shows the gut microbiota regulates intestinal T cell development by directing vitamin A flux through a sequential pathway involving epithelial, myeloid, and T cells.

This study identifies microbiota-derived short-chain fatty acids as key mediators of Candida albicans colonization resistance in the gut.

This study identifies an IL-17-DUOX2 axis that regulates C. albicans colonization and filamentation in the gut through interactions with candidalysin toxin.

This study identifies TDCA, a microbiota-derived bile acid, as a mediator that restores immune checkpoint therapy efficacy after antibiotic-induced hyporesponsiveness by enhancing CD8+ T cell antitumor immunity.

This study shows that the time of day affects cancer immunotherapy efficacy through a circadian clock-dependent chemokine axis in mice.

This study evaluates machine learning algorithms and feature selection stability in multi-omics microbiome datasets, finding that nonlinear models and multi-omics integration improve predictive performance and reproducibility.

IsoAnalytics provides a user-friendly web server for analyzing and visualizing single-cell proteomics data from the Isoplexis platform, accessible to non-experts.

MB-SupCon integrates microbiome and metabolomics data to improve prediction accuracy in microbiome-only datasets, achieving high accuracy in predicting insulin resistance, sex, and race.

MetaPrism enables joint analysis of taxa and gene abundances in metagenomic data to identify potential biomarkers for therapeutic response prediction.

The study shows that the TolC protein in E. coli is highly evolvable, making it a target for resistance against bacteriophages and bacterial toxins.

HARMONIES infers sparse microbiome networks using a zero-inflated negative binomial model and regularization, outperforming existing methods and identifying disease-associated microbial communities.

VAMPr predicts antibiotic resistance using genomic data by analyzing variant associations and building machine learning models.