Update: GDPR compliance - Klebo/eventflo Conferences
Poster
140

Pilot study to compare a traditional HTS approach with AI-driven compound selection for the identification of kinase inhibitors

Authors

T Chowdhury1; X Kodji1; S Maidment1; O Lazari1; I Lemasson1; J Ferdinand1; R Clark1; A Jones1; G Wishart1; H Van de Poel1; E Stangeland2; P Boardman2
1 Charles River Laboratories, UK;  2 Valo Health, United States

Discussion

Authors

T Chowdhury1; X Kodji1; S Maidment1; O Lazari1; I Lemasson1; J Ferdinand1; R Clark1; A Jones1; G Wishart1; H Van de Poel1; E Stangeland2; P Boardman2
1 Charles River Laboratories, UK;  2 Valo Health, United States

Discussion

HTS typically involves testing thousands of compounds in a step-wise manner, progressing putative hits from the initial single point analysis through to XC50 determination, with an array of hit confirmation and counter screen assays included in the screening cascade. Artificial Intelligence (AI) offers an alternative to the traditional HTS, with the potential to reduce the number of compounds being tested and thereby shorten the time to advance hits into later phases of drug development.

In this pilot study, AI modelling was performed by Valo Health, Charles River’s partner in transformative drug design, to select compounds for screening in parallel to a traditional HTS designed to identify inhibitors of a kinase implicated in inflammatory conditions.

ADP-GloTM technology was used to develop and validate an assay to screen 160,000 Charles River Lead-like and SoftFocus Kinase library compounds in singlicate at 12.5 µM in a 1536-well format. Approximately 650 putative hits from the HTS and 150 AI-selected compounds were progressed to concentration-response testing to determine the IC50. To eliminate non-specific compounds, a technology counter-screen was included to identify compounds interfering with the detection reagents. Sub-micromolar hits were identified with several of the AI selected compounds generating low nanomolar IC50’s. Early SAR was identified, and medicinal chemistry optimisation has been initiated.

This pilot study highlights the benefits of the two approaches employed for the identification of inhibitors of the target kinase, with the condensed HTS and focussed AI compound screening forming a powerful combination for the delivery of starting points for drug development. The integration of experimental and AI driven drug discovery can be accessed through LogicaTM, Charles River and Valo Health’s combined offering to rapidly deliver optimized preclinical assets.