Core Technology
Innovative drug discovery powered by nature and AI.
We harness proprietary AI to rapidly screen billions of plant-derived chemicals for promising bioactive properties — turning natural complexity into ranked, actionable candidates for our preclinical programs.
Key metrics
- Compound library
- 2B+
- Natural compounds
- 10M+
- Higher hit rate
- 64x
- Bioactivity prediction efficiency
- 44.83%
Virtually screenable molecules accessible through our AI pipeline.
Plant-derived bioactives indexed for screening.
Improvement over the most popular molecular docking software.
Increase reported in our IEEE DTIGN paper versus baseline.
Small Molecule Discovery Pipeline
Our pipeline is built upon three core modules. This holistic approach allows us to efficiently uncover and develop potent, nature-based compounds, accelerating the journey from discovery to market-ready treatments.
01
Botany and Mycology Genomics & Chemistry
ChEMBL Library Integration
02
AI and Computational Biology
Molecular Dynamic Simulations
03
Biomedical Validation
Mechanistic and ADMET Studies
Drug-Target Interaction Graph Neural Network
DTIGN — graph-based bioactivity prediction.
Our DTIGN model is the cornerstone of our AI-driven drug discovery platform. Engineered to accelerate and enhance the prediction of drug-target interactions, DTIGN plays a pivotal role in understanding bioactivity and driving groundbreaking innovations in drug development.
- · Enhanced hit selection and lead optimisation
- · Accelerated drug discovery timelines
- · Exploration of new therapeutic opportunities
- · Integrates Graph Neural Networks (GNNs), self-attention, and semi-supervised learning
Benchmark
27.03% better than other leading methods
When compared to other leading approaches, DTIGN performs 27.03% better at accurately predicting how strong a drug's effect will be on its target. The model uses pIC₅₀ and pEC₅₀ measurements alongside basic physics priors, learning from a small set of real molecular structures.
Publications
DTIGN: Advancing Bioactivity Prediction Through Molecular Docking and Self-Attention
Nanyang Biologics Research · IEEE Xplore, 2024
Introducing LigoSpace: Advancing Bioactivity Prediction at NeurIPS 2025
Nanyang Biologics in collaboration with Nanyang Technological University Singapore (NTU) · NeurIPS 2025, 2025
Research Briefs
Subscribe to NYB Research Briefs
Monthly digest of our AI drug discovery progress, publications, and longevity science.