Research Directions

Our laboratory develops computational and AI-based approaches to address challenging problems in biology and biomedicine. We combine machine learning, statistical modeling, and large-scale omics analysis to extract biological insights from complex data and to develop new approaches for understanding biological systems.

Our research covers antimicrobial resistance, epigenetic regulation, protein function annotation, drug discovery, and microbiome science. By bringing together computational methods and biological knowledge, we aim to develop reliable and interpretable tools that can help answer important biological questions and support biomedical research.

1. Antimicrobial Peptide Prediction and Design

Antimicrobial peptides (AMPs) are an important part of innate immunity and have attracted increasing interest as potential alternatives to conventional antibiotics. Our research focuses on computational methods for predicting antimicrobial activity and understanding the sequence and structural features that determine peptide function. We use machine learning and deep learning to evaluate peptide candidates, analyze their physicochemical properties, and guide the design of new sequences. We are particularly interested in identifying peptides with broad antimicrobial activity and favorable toxicity profiles.

2. DNA/RNA/Protein Modification Site Prediction

Chemical modifications of DNA, RNA, and proteins are closely involved in gene regulation, cellular signaling, and disease processes. Our research develops computational methods for identifying modification sites and characterizing the sequence, structural, and cellular context associated with these modifications. We combine sequence analysis, structural information, and multi-omics data to study how modification patterns vary across biological conditions. These approaches are applied to questions in epigenetic regulation, cancer biology, immune response, and precision medicine.

3. Protein Function Prediction

Protein function prediction is a fundamental problem in computational biology, particularly for proteins with limited or incomplete functional annotation. Our research focuses on computational approaches for inferring protein function from sequence, evolutionary, structural, and interaction information. We develop and evaluate models for functional annotation, biochemical activity prediction, and the characterization of proteins with previously unknown functions. This work provides computational support for studying protein function and molecular mechanisms in areas including disease biology and synthetic biology.

4. Small-Molecule Drug Development

Small-molecule drug development involves identifying compounds with suitable biological activity and understanding how they interact with their molecular targets. Our research uses computational modeling and machine learning to study compound–target interactions, predict molecular properties, and prioritize promising compounds for further investigation. We are interested in computational approaches to virtual screening, drug–target interaction prediction, molecular optimization, and drug synergy analysis, with a focus on improving the efficiency of early-stage drug discovery.