Principal Investigator (PI)
The Foundation is providing £208.760 in support.
Tiago Borges Janela
GSK R&D will serve as the experimental validation partner. The project will focus on commercially-available libraries and publicly available datasets. The outputs of the screening campaign will then be experimentally validated in the industrial setting provided by GSK. We believe this approach will offer compelling evidence for the effectiveness of open-source AI/ML platforms in anti-infective drug discovery.
We anticipate nominating 200 and 100 prioritized compounds for experimental validation at GSK, with a goal of confirming at least 5 active hits. This phase is critical to demonstrating the real-world applicability of the overall pipeline.
Artificial intelligence (AI) and machine learning (ML) hold promise for increasing the throughput and effectiveness of antimalarial hit identification. However, adoption of AI/ML in antiparasitic research laboratories remains slow, largely because of data scarcity and the technical burden of developing and deploying these tools. This challenge is particularly acute in laboratories in the Global South, where computational infrastructure and programming expertise are often insufficient to support truly AI/ML-driven research.
To address these barriers, we are developing the Ersilia Model Hub, the largest open-source platform of AI/ML models tailored to the anti-infective drug discovery community, with a focus on malaria, tuberculosis, and antimicrobial resistance, among others.
In this project, we aim to develop the most comprehensive ensemble of predictive models for antiplasmodial activity, incorporating both phenotype-based and target-based approaches. This array of AI/ML bioactivity predictors will be used to perform a billion-scale virtual screen of commercial (make-on-demand) libraries, with the goal of identifying novel, potent chemotypes active against Plasmodium spp. All AI/ML models and screening results will be made publicly available. Selected virtual hits will be procured and experimentally validated at GSK-TCOLF.
The project will deliver two primary outputs:
- A set of experimentally validated hits against Plasmodium spp.
- A collection of production-ready AI/ML models for antimalarial drug discovery, released openly to the global scientific community.
Together, these outputs will accelerate the discovery of new antimalarial therapeutics while making AI/ML-driven approaches more accessible to laboratories worldwide.