Microbiome
foundation models
Learning transferable representations from large-scale microbial data to support diverse biological questions.
Bioinformatics · AI for biology
Learning the language
of microbial life.
I build foundation models to understand microbial communities and connect biological complexity with interpretable AI.
A unified foundation model for microbiome world exploration.
bioRxiv · PreprintI am a direct-track Ph.D. student in bioinformatics and computational biology at HUST, working in Prof. Kang Ning’s group.
My research brings deep learning and large-scale pretraining to microbiome analysis. I am interested in representations that transfer across communities, cohorts and biological contexts—and in making what these models learn interpretable.
My work spans microbiome foundation models, microbial community dynamics, biosynthetic gene clusters and antimicrobial peptide discovery.
ORCID 0000-0001-6267-4244Direct-track · Huazhong University of Science and Technology
Huazhong University of Science and Technology
Generalizable models.
Interpretable representations.
Biologically meaningful insights.
Learning transferable representations from large-scale microbial data to support diverse biological questions.
Modeling how microbial communities change over time, with context-aware prediction and digital-twin frameworks.
Exploring biosynthetic gene clusters and antimicrobial peptides with context-aware protein language models.
Transferring knowledge across regions and datasets for robust microbiome-based classification and prediction.
A unified foundation model for microbiome representations, prediction and interpretation.
View repositoryLarge-scale pretraining for transferable microbiome analysis across diverse contexts.
View repositoryA genome foundation model that learns discrete, context-aware genome tokens from ordered prokaryotic protein sequences.
View repositoryContext-aware microbial source tracking with transfer learning and biome ontology.
View repositoryMicrobiome-based models for the diagnosis of a broad spectrum of cancer types.
View repositoryInterpretable survival subtyping from microbial profiles and host gene expression.
View repository