Multi-Omics Network Analysis

We develop novel network science methods and graphlet-based algorithms to analyze multi-omics networks at multiple scales. By utilizing local network topology, we correct for information missed by traditional random walks and enable embedding of multi-omics data into more linearly separable spaces. This allows for explainable and sustainable mining of complex biological networks to understand the molecular organization of life and disease.

Multi-omics data fusion AI for precision medicine

We develop versatile artificial intelligence frameworks for multi-omics data fusion, based on graph-regularized non-negative matrix tri-factorization (NMTF) and other machine learning techniques. Our methods enable dimensionality reduction, inference, and co-clustering of heterogeneous datasets to better stratify patients into risk groups, discover new biomarkers and therapeutic targets, and to re-purpose existing and discover new drugs for personalized medical treatment.

Precision Therapeutics Discovery

We apply our computational frameworks to address key challenges in precision medicine across diverse diseases. From time-series and patient-derived single-cell data and all other multi-omic data, we develop consumer-facing precision medicine solutions for cancer, infectious diseases (including Covid-19 and antimicrobial resistance), neurodegenerative disorders (Parkinson’s), and longevity research. Our goal is to translate multi-omics insights into actionable therapeutic discoveries, including novel antibiotics, anti-cancer and anti-aging compounds, that improve patient outcomes.