Research Overview

Decoding Microbial Life with AI & Metagenomics

My research lies at the intersection of computational biology, machine learning, and microbiology. I develop and apply deep learning models to explore massive metagenomic datasets, with the goals of describing planetary viral diversity and elucidating functions of novel genes.

Virology Functional Annotation Microbiome Science

Core Research Directions

Biological Sequence Analysis

  • CRISPR-PAM: Metagenomic mapping of PAM preferences across millions of Cas9 proteins to expand targeting options.
  • DeepFRI: GNN-based function prediction directly from sequence-derived contact maps at database scale.
  • scikit-bio: Contributions to Python's core bioinformatic library for sequence analysis and diversity metrics.

Microbiome - Host Relationship

  • Probiotics: Clinical trial mapping multi-species probiotics response on gut microbiota functions in obese postmenopausal women.
  • Rhinosinusitis: Investigating spatial variability in chronic rhinosinusitis sinus microbiome, highlighting patient-specific differences.

Machine Learning

  • Metagenomic DeepFRI: Integrating deep-learning GNNs into bioinformatics pipelines to annotate novel gut microbial genes.
  • TM-Vec2: Lightweight sequence embedder trained via knowledge distillation to predict protein structure similarity (TM-scores) at scale.