Mining the First Fruits of AI Drug Discovery from Patents
Year: 2025
Type: article
Abstract: After “talking the talk” AI-centric drug discovery companies (AIDDCs) are now “walking the walk” by not only progressing lead molecules to clinical trials but also disclosing their AI-designed novel chemical series for drug targets in published patents. This presentation outlines mining these documents to provide a valuable snapshot of what data AIDDCs are actually generating, extracting SAR data sets from BindingDB and PubChemBioAssay and analysing these in CDD Vault. to give unique insights. Examples will be shown of data and chemistry extractions via open tools and sources from patents published by Exscientia, BenevolentAI, Atomwise, Insilico Medicine. This presentation will focus on how the continually developing CDD Vault toolbox can be used to upload patent-extracted quantitative data sets for the analysis of chemistry and SAR data coming into PubChem BioAssay via BindingDB(BDB) curation. The demonstrated functionality includes various plots of LogP, Mw and IC50 values as well as novelty checking searches of specific lead compound candidates launched from within CDD Vault. (Slides presented at the August 2025 ACS meeting are also available at //zenodo.org/records/17020396)
Author: Southan, Christopher
Institution: University of Edinburgh
Language: English
Subfield: Computational Theory and Mathematics
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7082375939
FWCI:
Cites:
Cited by:
Related to: