Yuzhen Ye

The paper benchmarks toxicity prediction models and introduces a 3D-structure-based deep learning model that improves classification accuracy.

FIDDLE is a deep learning method that predicts chemical formulas from tandem mass spectra, achieving high accuracy and speed compared to existing methods.

This study develops confounder-free predictive models for host phenotype prediction using microbiome data, showing improved accuracy and biological interpretability.

The paper introduces MicroKPNN-MT, a model that predicts human phenotypes and missing metadata from microbiome data, improving disease prediction accuracy and model generalizability.

This paper introduces a panproteome graph to enhance peptide and protein identification in proteomics and metaproteomics by leveraging the human pangenome.

The paper introduces DCTdomain, a method that segments proteins into domains and uses discrete cosine transformation to create domain-level embeddings for improved protein similarity search.

The paper introduces MicroKPNN, a microbiome-based predictive model that improves host phenotype prediction by incorporating metabolic activity, taxonomy, and community structure, outperforming existing methods in multiple diseases.

3DMolMS predicts tandem mass spectra from 3D molecular conformations, achieving high cosine similarity with experimental spectra and enabling generalized predictions across instruments.

CAMMiQ improves microbial strain-level detection and quantification in single-cell metagenomics by using combinatorial optimization and doubly-unique substrings, outperforming existing methods in accuracy.

The paper performs a meta-analysis of microbial association networks to identify dysbiosis patterns in diseased gut microbiomes, revealing network features that distinguish healthy from disease states.

The paper analyzes CRISPR-Cas systems in Bacteroides fragilis genomes to understand their dynamics and adaptation in the human gut, revealing stable and dynamic spacer acquisition patterns and host-MGE interactions.

The paper uses metaproteomics to analyze human gut microbiome data, identifying generalist and specialist bacterial species and recovering ORFs missed by gene predictors.

This study introduces myRT, a tool for identifying and classifying reverse transcriptases in bacterial genomes and metagenomes, revealing DGR-related RTs are more prevalent in gut metagenomes.

The paper introduces HAPiID, a two-step method for faster and more effective peptide/protein identification in human gut metaproteomic data by using high abundance proteins as guides.

The paper develops a method to quantify metabolic interactions between microbial species while adjusting for phylogenetic relationships, revealing niche differentiation and habitat filtering in gut microbiomes.

The paper constructs a phylogenetic tree of human gut bacterial species using over 3000 genomes, including MAGs, and applies it to metagenomics and metaproteomics data analysis.

The paper develops CRISPRtrack, a pipeline using CRISPR spacer content to track donor strains in fecal microbiota transplantation and quantify microbiome similarity.