Davis Winkler has an unusually broad formal training in chemistry, physics, chemical engineering, and radioastronomy. He is a Professor of Biochemistry & Chemistry at La Trobe Institute for Molecular Science at La Trobe University, an adjunct Professor of Medicinal Chemistry at the Monash Institute for Pharmaceutical Sciences, and a visiting Professor in Pharmacy at the University of Nottingham. He previously spent over 30 years at CSIRO applying computational chemistry, AI, and machine learning methods to the design of drugs, agrochemicals, nanomaterials, and biomaterials. He is ranked 136th out of 100,000 medicinal chemists (Stanford 2023). He has authored over 250 refereed journal articles and book chapters (6 that are ISI Highly Cited), has an H index of 62 (GS), and is an inventor on 25 filed patents. He has provided key IP for three biotech startup companies. His awards include the CSIRO Medal for Business Excellence, RACI’s Adrien Albert award for medicinal chemistry and a Distinguished Fellowship, the ACS Herman Skolnik award for excellence in cheminformatics, a Royal Academy of Engineering (UK) Distinguished Fellowship (bioengineering) and the AMMA Medal (molecular design). He is past President of the Federation of Asian Chemical Societies (FACS) and the Asian Federation for Medicinal Chemistry (AFMC), and past Chairman and Director of the RACI Board
Topic of Presentation: The Exciting Potential of AI for Drug and Therapeutic Discovery
David A. Winkler
Professor of Biochemistry & Chemistry, La Trobe Institute for Molecular Science, La Trobe University, Bundoora 3086, Australia; Professor of Medicinal Chemistry, Monash Institute of Pharmaceutical Sciences, Monash University, Parkville 3052, Australia; Professor of Pharmacy, University of Nottingham, Nottingham NG7 2RD, UK
D.winkler@latrobe.edu.au
The current era has seen truly paradigm shifting scientific developments. We also know that the size of small molecule and materials spaces is for all practical purposes infinite, providing an inexhaustible supply of potential drugs and materials with valuable properties if we can find them. This recognition has seen a rapid increase in automation and robotics, allowing synthesis of new molecules and materials and measurement of properties orders of magnitude faster than before, generating massive databases of complex genetic, structural, chemical, property, and biological information.
The need to find ‘islands of chemical utility’ in an almost infinite palette of possibilities has seen the development of new data-driven machine learning methods (ML), like deep learning and large language models (LLMs), and an unprecedented increase in their applications. ML algorithms are universal approximators, prompting a parallel rise in their applications to most other aspects of modern life – medicine, finance, manufacturing, social media etc. we now have rapid and accurate quantum machine learning methods, generative methods that use trained machine learning models to suggest new molecules or materials with improved properties, accurate prediction of protein structures from sequence using AlphaFold and similar, the beginnings of general AI in LLMs like ChatGPT, massive ‘make on demand’ chemical libraries such as ZINC-22, nascent autonomous chemical discovery systems, and a rise in evolutionary AI methods for molecule and materials discovery.
This paper will discuss the drivers for these exciting developments and give examples where my collaborators and I have used AI and machine learning in biomaterials and regenerative medicine, drug design, energy materials and nanomaterials, cancer diagnostics, and corrosion control.
Keywords: Machine learning; drug design; artificial intelligence, materials discovery
?References
1. Hart, et al. Trust Not Verify? The Critical Need for Data Curation Standards in Materials Informatics. Chem. Mater. 2024 36, 9046
2. Winkler, Computational repurposing of drugs for viral diseases and current and future pandemics, J. Math. Chem. 2024, 62, 2844
3. Li et al. Rational Atom Substitution to Obtain Efficient, Lead-Free Photocatalytic Perovskites Assisted by Machine Learning and DFT Calculations, Angew. Chem. Int. Ed. 2023, e202315002
4. Thamarai Krishnan et al. Staging of colorectal cancer using lipid biomarkers and machine learning, Metabolom., 2023, 19, 84
5. Machine Learning for Electrocatalyst and Photocatalyst Design and Discovery, Chem. Rev. 2022 122, 13478
6. Muratov et al. A critical overview of computational approaches employed for COVID-19 drug discovery. Chem. Soc. Rev., 2021, 50, 9121
7. Piplani et al. In silico comparison of SARS-CoV-2 virus spike protein-ACE2 binding affinities across species; significance for animal susceptibility and viral origin, npj Sci Rep., 2021, 11, 13063
8. Szabo et al. Potent peptide antagonists of the thrombopoietin receptor as potential myelofibrosis drugs. Adv. Therapeut. 2021, 2000241
9. Muratov et al. QSAR without Borders, Chem. Soc. Rev. 2020, 49, 3525
10. Discovery and optimization of materials using evolutionary approaches. Chem. Rev. 2016,116, 6107
11. Autefage et al. Sparse feature selection methods identify unexpected global cellular response to strontium-containing materials. Proc. Natl. Acad. Sci. USA 2015, 112, 4280