This company uses proprietary AI technology to rapidly discover novel drug candidates against any protein target, even those without known structures. Their platform screens billions of molecules in under a day to accelerate hit identification and explore broad chemical spaces. They also offer hit refinement services by integrating proprietary data to optimize existing leads through fast prediction and testing cycles.
Funding
Funding not disclosed
Founders
Product
Problem
Traditional drug discovery methods for identifying small molecule hits often require known protein structures, which limits the druggable target space and increases R&D costs. Structure-based virtual screening can be computationally intensive and may not effectively explore ultra-large chemical spaces, hindering the discovery of novel drug candidates for challenging targets.
Solution
Factorize.bio offers an AI-driven platform that enables ultra-fast hit identification for any protein, even those with unknown structures. The platform leverages protein sequences and a proprietary AI model trained on over 80 million protein-ligand activity data points to predict binding affinities. By screening over 10 billion molecules in less than a day, Factorize.bio accelerates drug discovery, unlocks previously inaccessible targets, and facilitates the exploration of unprecedented chemical spaces. The technology bypasses the need for complex structural data, using SMILES representations for molecules and amino acid sequences for proteins, enabling confident predictions across diverse targets.
Target Audience
The primary customers are pharmaceutical companies, biotech firms, and research institutions seeking to accelerate early-stage drug discovery and identify novel drug candidates for challenging or previously undruggable protein targets.
Features
- AI-driven hit identification using protein sequences, eliminating the need for 3D structures
- Virtual screening of over 10 billion ready-to-order molecules per day
- Integration of proprietary data to personalize predictions and align with specific therapeutic objectives
- Rapid make/test/refine cycles for hit-to-lead development
- Optional add-ons including property-based filtering, selectivity and off-target insights, toxicity risk assessment, and predicted binding residues
- Ready-to-order SMILES list containing the predicted hits with model scores, prediction confidence, and refinement statistics