Explore uncharted chemical space
Chemistry42 is an automated machine learning platform for drug design capable of finding novel lead-like structures in hours
Automated de-novo drug designOperate beyond existing screening libraries and skip the effort of scaffold search and structure optimization. Chemistry42 is a fully-automated machine learning platform that delivers new lead-like structures in days
Scalable engineering platformChemistry42 is a seamlessly scalable distributed platform that can be deployed in cloud and on-premise environments with predictable hardware-agnostic workload management
Core Features
Annotate the generated molecules with their predicted ADME-PK and kinase selectivity profiles
Customizable Reward Function via API
Drive the design of small molecules with the right profile based on the reward settings
Tailored Reward Function
Discover novel lead-like structures in days
Generative AI-driven Active Learning Platform
Version 2.1-2.2 Updates
May 2023
Improved Performance: Generate more diverse novel chemotypes that are synthetically accessible
Updated Novelty Dataset: Inclusion of ChEMBL31, resulting in the addition of 2,200,048 unique compounds
New Drug-likeness Properties: Control the flexibility, number of tetracycles and bicycles and more
Updated Novelty Dataset: Inclusion of ChEMBL31, resulting in the addition of 2,200,048 unique compounds
New Drug-likeness Properties: Control the flexibility, number of tetracycles and bicycles and more
extensive experience in artificial intelligence AND MEDICINAL CHEMISTRY
3 Simple steps to get
Actionable lead-like molecules
Actionable lead-like molecules
Artificial intelligence can design novel molecules for any available target structure or small molecule ligand, creating and testing novel scaffolds across the desired properties and optimizing binding affinity
Ligand/Structure-based
Drug Design
Drug Design
Define rewarding and penalty rules for molecule shape, chemical complexity, synthetic accessibility, metabolic stability, and other properties the novel molecules must satisfy
Guide the generation
Every new compound generated is annotated with all the properties, including physico-chemical parameters, binding scores, drug-likeness features and mapped on vendors' catalogs and proprietary libraries for any similarity and novelty
Molecular annotation
COLLABORATIONS