Alvin Chan Lab AI for Therapeutic Discovery

AI for Therapeutic Discovery

We are an interdisciplinary research group at Nanyang Technological University, jointly based in the College of Computing and Data Science and the Lee Kong Chian School of Medicine. We integrate artificial intelligence and high-throughput experimental science to discover safer, more effective therapeutics for intractable diseases.

Our Research

AI-driven therapeutic discovery

AI-driven therapeutic discovery

Our research covers small molecules, proteins, peptides, RNAs, and LNP formulations. Across these domains, we develop AI methods for three core tasks: predicting properties, interactions, and biological outcomes; designing new therapeutic candidates; and optimizing candidates under multiple objectives.

Closed-loop therapeutic discovery

Closed-loop therapeutic discovery

We develop closed-loop discovery systems that connect AI-guided candidate development with computational and experimental evaluation. Predictive models assess candidate properties and outcomes, generative models propose new candidates, and optimization methods refine them across multiple objectives. Candidates are then evaluated through computational screening, simulation, experimental assays, and expert feedback. The resulting measurements and feedback are converted into learning signals that update the models and guide the next round of discovery.

People

One interdisciplinary lab

One interdisciplinary lab

Researchers in AI and experimental science work together on shared biomedical problems, connecting model development with experimental evaluation and real therapeutic applications.

Selected Publications

Towards Understanding Modality Interaction in Multimodal Language Models via Partial Information Decomposition
Towards Understanding Modality Interaction in Multimodal Language Models via Partial Information Decomposition
Wanlong Fang, Tianle Zhang, Wen Tao, Alvin Chan
Forty-Third International Conference on Machine Learning (ICML 2026)  ·  2026
Understanding modality interaction in multimodal large language models (MLLMs) is central to reliable deployment. We introduce Partial Information Decomposition (PID) as a decision-level framework that separates unique, redundant, and synergistic contributions of sensory and linguistic inputs, beyond representation alignment and outcome-based evaluation.
Designing lipid nanoparticles using a transformer-based neural network
Designing lipid nanoparticles using a transformer-based neural network
Alvin Chan, Ameya R. Kirtane, Qing Rui Qu, Xisha Huang, Jonathan Woo, …, Miguel Jimenez, Michelle Sun, Yuebin Huang, Ceara Byrne, Giovanni Traverso
Nature Nanotechnology  ·  2025
The RNA medicine revolution has been spurred by lipid nanoparticles (LNPs). The effectiveness of an LNP is determined by its lipid components and their ratios; however, experimental optimization is laborious and does not explore the full design space.