The invaluable role of machine learning methods at different stages of drug design pipelines
Computer-aided methods are nowadays an essential component of the drug discovery workflow, providing useful tools various stages of this expensive and time-consuming process. Their service starts at the very beginning, when computational tools are applied to identify new compounds with potentially desirable biological profiles. However, in silico evaluation is not only limited to the assessment of the activities of the investigated molecules towards considered targets but also involves analysis of their physicochemical and pharmacokinetic properties and potential toxicities, as well as generation of new potential drug candidates.
The presentation will summarize the areas of application of machine learning methods in projects implemented in the Department of Medicinal Chemistry Maj Institute of Pharmacology Polish Academy of Sciences. They involve both methodological initiatives aimed at the development of novel machine-learning-based algorithms, as well as projects oriented at the search of structurally novel ligands of selected G protein-coupled receptors (mostly serotonin and opioid receptors). In addition, the on-line platforms allowing the usage of the constructed tools by wide community will be presented. They refer mostly to the prediction of compound ADMET properties and indication of chemical moieties influencing to the highest extent the value of the considered parameter.
I am currently Assistant Professor in the Department of Medicinal Chemistry Maj Institute of Pharmacology Polish Academy of Sciences, where I have worked since 2010. I obtained my PhD (discipline: chemistry) in 2016 and earned habilitation in 2023 (discipline: pharmaceutical sciences).
After completing my PhD, I was head of my three-year post-doc grant SONATINA, which I run at the Faculty of Pharmacy Jagiellonian University Medical College and currently I am principal investigator of the grant OPUS, which is implemented in IP PAS. So far, I have been PI in five research grants funded by the National Science Centre, and I have been member of the research team in 13 another projects. My research work resulted in 47 articles in scientific journals, 1 book chapter, and over 70 conference presentations. I was also appreciated by the Foundation for Polish Science and awarded the START scholarship in 2019.
I am chemist and mathematician by education and in my work I focus on the application of machine learning methods in the process of searching for new drugs with the particular attention paid to the ligands of the G protein-coupled receptors. Recently, the core of my work are explainability approaches and on-line applications of the developed predictive tools.