We apply state-of-the-art AI to high-impact scientific domains — advancing molecular discovery in bioinformatics, building foundation models for time series analysis, and accelerating functional material design for real-world applications.
We apply state-of-the-art AI to high-impact scientific domains — advancing discovery in molecular biology, sequential data analysis, and functional materials design.
We apply state-of-the-art generative AI and deep learning to molecular-level bioinformatics problems, spanning ligand-based and structure-based drug discovery, scaffold-based molecular design, binding affinity prediction, and explainable models for biological insight.
We investigate fundamental and applied AI methods for time series and sequential data, developing foundation models, retrieval-augmented imputation systems, visual forecasting frameworks, and multimodal analytical pipelines that handle real-world temporal complexity.
We use AI and computational modeling to discover and design functional materials with targeted real-world properties, focusing on thermal-management materials for agricultural environments, stealth and camouflage materials for defense applications, and smart materials for next-generation biosensor platforms.
From generative molecular design to foundation models for time series — explore the projects shaping our research agenda.
Bioinformatics
A diffusion-based generative framework for de novo molecular design, enabling scaffold-aware generation of drug-like molecules with controllable structural and pharmacological properties.
Bioinformatics
A structure-informed deep learning model for protein–ligand binding affinity prediction, combining 3D geometric features with sequence-level representations for cross-target generalization.
Time Series
A retrieval-augmented generation approach for time series imputation that retrieves semantically similar historical patterns to guide missing value recovery in irregular multivariate signals.
Materials
AI-guided design of thermally adaptive composite materials for passive cooling of agricultural greenhouses, reducing peak internal temperature without active energy consumption.
Materials
Generative models for spectrally selective surface coatings providing broadband camouflage across visible, near-infrared, and thermal infrared spectra for defense vehicle applications.
Peer-reviewed research published at international journals and conferences spanning AI for bioinformatics, time series, and language models.
Tan Khanh Nguyen*, Thi-Thu Nguyen*, Khanh Huyen Thi Pham, Manh-Tu Luong, Nhat-Hai Nguyen
Thi-Thu Nguyen*, Xuan-Thong Truong*, Khac-Thai-Binh Nguyen, Nhat-Hai Nguyen
Thi-Thu Nguyen, Nhat-Hai Nguyen, Quoc-Tuan Ngo
Research updates, technical deep-dives, and perspectives from our team across bioinformatics, time series, and materials.
We review recent progress in applying score-based and DDPM-style diffusion models to molecular generation, and outline the open challenges in conditional, scaffold-aware, and target-specific generation.
What can time series researchers learn from the success of large language models, and what makes sequential numerical data fundamentally different from text? We discuss our experiences building TSFounder.
How our ThermoShield project uses generative models and physics-informed optimization to identify novel composite materials that passively reduce greenhouse heat load by up to 18°C.
Collaborating with world-class institutions
We welcome PhD candidates, postdoctoral researchers, and industry collaborators passionate about applying AI to bioinformatics, time series, and materials science.
We offer competitive stipends, world-class mentorship, and access to cutting-edge compute infrastructure.