Accelerated Discovery and Development of New Medicine

Professor David Palmer and Huw J. Williams, in the Department of Pure and Applied Chemistry at the University of Strathclyde, are using ARCHIE-WeSt to contribute to the advancing field of property prediction for pharmaceutical candidates. Early-stage drug discovery is increasingly reliant on computational and data-driven approaches to molecular design and selection; however, there remains a need for further optimisation of these methods to improve the integration of computation into drug discovery workflows.

To date, this work has focused on the evaluation of query strategies within early-stage pharmaceutical drug discovery processes [J. Chem. Inf. Model. 2026, in press]. By simulating the Design–Make–Test–Analyse (DMTA) cycle using computational methods such as molecular docking and machine-learning-based property prediction, we can assess the long-term effects of different query strategies on models used to propose future generations of synthetic candidates, helping to probe the gap between computationally and empirically valuable information.
Alongside this methodological development, the project has contributed to the identification of real hit candidates through the GSK Prosperity Partnership. In this work, deep learning frameworks for property prediction have been used to guide molecule selection for synthesis and subsequent assay testing, with the aim of identifying novel hit compounds targeting bromodomain-containing proteins.
Future work will focus on the development of improved molecular representations for predictive models, with particular emphasis on evaluating how effectively molecular encoders abstract key information relative to traditional molecular descriptor packages.

For a list of the research areas in which ARCHIE-WeSt users are active please click here.