Development of Advanced Machine Learning and Artificial Intelligence Models in an Open Cloud-Based Platform to Accelerate Drug Discovery (AIRCHECK)

The Artificial Intelligence Ready CHEmiCal Knowledgebase (AIRCHECK) is an open platform developed to share and analyze large-scale, freely accessible chemical activity data. AIRCHECK is specifically designed to integrate high-throughput chemical screens data generated from DNA-encoded libraries (DEL) and Affinity Selection Mass-Spectrometry (ASMS). AIRCHECK aims to become a global platform that advances Machine Learning (ML) and artificial intelligence (AI) in drug discovery through open collaboration across academia and industry. The goal is to create a dynamic, inclusive, and collaborative community that accelerates innovation in chemical biology through the availability of analysis-ready datasets and open-source AI models that are transparent, reproducible, and reusable. To support researchers worldwide, AIRCHECK aims to offer resources (data and computation) and host workshops, webinars, and competitions to become a cornerstone for collaborative drug discovery and a way to open science in chemical biology.
The Structural Genomics Consortium (SGC) supports the project by contributing its expertise in data curation, standardization, and fostering open science collaborations between academia and industry. SGC plays a key role in ensuring the quality and accessibility of the datasets integrated into AIRCHECK.
This proposal seeks to enhance AIRCHECK by improving data management strategies, standardizing DEL and ASMS datasets, and developing machine learning models with benchmark datasets. We will focus on building novel algorithms for large-scale chemical screening and a Machine Learning Operations (MLOps) framework to streamline data ingestion, model development, testing, and deployment, enabling collaborators to focus on advancing ML/AI techniques without the complexities of managing the research lifecycle.

Faculty Supervisor:

Benjamin Haibe-Kains

Student:

Partner:

Structural Genomics Consortium

Discipline:

Computer science

Sector:

Professional, scientific and technical services

University:

University of Toronto

Program: