
Drug Discovery Solutions (ALScreen, PELETac)
AI and molecular modeling for multi-stage drug discovery, from target selection and hit finding through lead optimization and targeted protein degradation.
Overview
Nostrum Biodiscovery's Drug Discovery Solutions — encompassing the proprietary ALScreen and PELETac platforms — deliver a custom-tailored, AI-powered computational service designed to accelerate drug discovery across multiple stages. Combining state-of-the-art artificial intelligence with molecular modelling (AI+MM), the platform supports projects spanning small molecule design, targeted protein degradation, vaccines, antibodies, and cosmetics applications, making it a comprehensive partner for pharmaceutical and biotech organisations at any point in their discovery pipeline.
The service is structured around the key phases of modern drug discovery, from initial target selection through to lead optimisation, with specialist capabilities in targeted protein degradation for challenging undruggable targets. Each stage is supported by integrated computational tools and expert guidance from experienced drug design scientists.
Target Selection
- Protein structure prediction for targets with unknown 3D structures
- AlphaFold-based conformer generation to incorporate conformational dynamics
- Druggability assessment of therapeutic targets
- Target binding site identification and validation
- Reverse docking to identify potential binding partners
- Cysteine accessibility assessment to support covalent inhibitor development
Hit Finding with ALScreen
- Proprietary ALScreen platform combining AI and molecular modelling techniques for cost- and time-efficient hit identification
- AI+MM-based virtual screening of predefined and ultra-large compound libraries
- Pharmacophore screening to identify structurally relevant candidates
- Covalent inhibitor screening
- Fragment-based screening approaches
Hit to Lead
- De novo generation of small molecules
- Similarity search across chemical space
- Fragment growing using Frag-PELE methodology
- Bioisostere replacement for scaffold diversification
- Free energy predictions to guide compound selection
- Pharmacophore modelling
- Target flexibility modelling
- Dual inhibitor design and polypharmacology strategies
- Covalent binder development
Lead Optimisation
- 3D machine learning models to guide structural refinement
- Rational design and medicinal chemistry support
- Free energy prediction for potency and selectivity optimisation
- In silico ADMET prediction to assess drug-like properties early
Targeted Protein Degradation (TPD)
- Custom-tailored development pipelines for both Molecular Glues and PROTAC discovery
- Molecular glue development for a given binary complex
- E3 ligase pairing for a specified protein of interest (POI)
- Binary and ternary complex prediction
- Proprietary E3 ligase-warhead library for PROTAC design
- PELETac platform employing a sequential approach combining protein-protein perturbations with PROTAC sampling to exhaustively map ternary complex formation, propensity relative to other PROTACs, and overall ligase complex geometry
Nostrum Biodiscovery's drug discovery capabilities are underpinned by a strong publication record in leading journals including Nature Communications and ACS Publications, covering topics such as AlphaFold2-based virtual screening, generative AI merged with active learning frameworks, free energy calculations, and genetic variability impacts on drug sensitivity. The platform is suitable for organisations seeking a scientifically rigorous, computationally driven partner to accelerate and de-risk their drug discovery programmes.
