Kang Ning

The paper introduces NK-RCPP, a workflow for prioritizing neurokinin-receptor peptide candidates using receptor-matched scores and structural-neighbour comparisons.

PaleoRigor uses a language model with constraints to process paleomicrobiome data, ensuring reproducibility and preventing invalid operations.

The paper introduces MGM2, a multimodal foundation model for microbial communities that predicts responses to interventions and interprets community changes using shared representations of molecular identity and context.

MetaClaw is an auditable AI agent that performs end-to-end metagenomic and multi-omics analysis, recovering key biological markers across multiple studies.

The paper introduces MGM, a foundation model for microbiome analysis, showing improved classification and detection of dynamic patterns across diverse contexts.

The paper develops MicroSPEAR, a framework combining a protein language model and microfluidics to discover and validate antimicrobial peptides from isolated habitats.

This study develops the Tumor Microbiome Survival Index (TMSI) to predict cancer patient survival and drug sensitivity across multiple cancer types based on intra-tumor microbiome data.

MicroProphet predicts microbial community dynamics from sparse longitudinal data, enabling personalized health monitoring and intervention planning.

BGC-Finder uses protein language models to detect and annotate biosynthetic gene clusters in microbial genomes and metagenomes, improving speed and detection of core genes and cytochrome P450s.

This paper introduces BGC-Prophet, a transformer-based model for identifying and extrapolating biosynthetic gene clusters (BGCs) in microbial genomes and metagenomes, revealing evolutionary patterns and enhancing BGC discovery.

This paper introduces MGM 2.0, a generative language model for microbiome analysis that uses sentence-like representations to predict microbial colonization, generate disease-specific profiles, and improve donor selection for fecal microbiota transplantation.

SMeta improves metagenome binning by integrating single-cell and metagenomic data to enhance accuracy in reconstructing microbial species.

The paper develops DeepInverton, a deep learning model to identify invertible promoters in bacterial genomes and metagenomes, revealing their prevalence and functional roles in microbial adaptation.

This study develops ASD-cancer, a deep learning framework that identifies survival subtypes in cancers by integrating tumor microbiome and transcriptome data, revealing microbial-gene interactions linked to patient outcomes.

The paper introduces Meta-Sorter, an AI method using neural networks and transfer learning to improve biome labeling for thousands of inaccurately annotated microbial samples, achieving high accuracy in classification.

This paper introduces ONN4MST, a deep learning method for microbial source tracking that achieves high accuracy and speed across large datasets of microbial communities.

This paper introduces ONN4ARG, an ontology-aware deep learning method for antibiotic resistance gene discovery, identifying 120,726 candidate ARGs, including novel ones not in public databases.

GAN-GMHI improves disease prediction by enhancing the discrimination power of a gut microbiome-based health index through batch effect removal using a generative adversarial network.

The paper improves protein structure prediction by linking microbiome sequences to protein families using a targeted approach with predicted biomes.

The paper introduces ContactLib-ATT, a two-level neural network for protein structure embedding that improves SCOP superfamily classification and remote homolog search.

EXPERT uses transfer learning to improve microbial source tracking, showing better efficiency and accuracy in tracking gut microbiome changes in infants and colorectal cancer patients.

This study evaluates the multi-barcoding approach for identifying herbal materials in TCM preparations using ITS2 and trnL biomarkers, showing high detection rates and improved reliability with combined use.

QC-Blind removes contaminants from bacterial DNA sequencing data without a reference genome, preserving target species genomic information with high specificity and accuracy.

The paper uses bi-clustering to identify gene clusters with consistent expression-protein abundance correlations and builds a prediction model to forecast protein abundance from mRNA data.

The paper develops a comprehensive ITS2 sequence database and search tools for plant species identification, providing a web service for accurate and efficient marker analysis.

The paper analyzes marine antibiotic resistance genes (ARGs) and their associations with bacterial communities using Tara Oceans metagenomic data, finding higher ARG abundance and diversity in mesopelagic zones and ARG-enriched genera being more abundant than their ARG-less counterparts.

This study examines how agricultural pollution affects microbial communities in Honghu Lake, finding that physicochemical and antibiotic factors drive changes in microbial diversity and structure.