ARTIFICIAL INTELLIGENCE AND THE DISCOVERY OF NEW MEDICINES
On Wednesday, April 24, 2024, at the headquarters of the OncoGen Center, a very interesting conference was held mediated by Prof. Dr. Virgil Păunescu, the coordinator of the center and supported by Prof. Dr. Tudor Ionel Oprea, Professor Emeritus at the Faculty of Medicine of the University of New Mexico, USA and Chief Scientific Officer at Expert Systems Inc, San Diego CA, USA.
Prof. dr. Tudor Oprea is a graduate of the "Victor Babeș" University of Medicine and Pharmacy in Timișoara and a collaborator of the OncoGen Center team for many years. We highlight the fact that Tudor Oprea is a global professional with scientific leadership and management skills in drug discovery and complex data integration, having decades of experience in developing computational methods and implementing computational models in drug research. Often presenting himself as a "digital drug hunter", Tudor Oprea brought together his multidisciplinary experience in fields that include medicine, chemistry, informatics, computational biology, machine learning, data science, becoming an expert in the repositioning of medicinal substances, chemoinformatics and QSAR (Quality Structure -Activity Relationship), virtual screening and identification of molecular targets.
Currently, the discovery of new drugs or the identification of new therapeutic targets for classic drugs requires the cooperation of life science researchers, programmers, clinicians, pharmacologists, chemists, regulatory experts. They operate in a complex context, facing multiple challenges in the field of health, but also benefiting from advances in information technology (artificial intelligence, machine learning, computational biology). At the same time, they have at their disposal huge amounts of biological and chemical data (genomic data, protein structures, chemical compounds) and personalized medicine approaches.
In his presentation, the guest emphasized the fact that artificial intelligence/machine learning (AI/ML) systems can make valuable contributions both to drug target reuse, selection and validation, identification of phenotypic associations, and to finding new therapeutic modalities and reuse/ drug repositioning using in silico methods.
Using multiple parallel IA/ML models helps highlight differences and similarities between diseases at the genomic level. This can translate into potential selectivity for certain diseases (different targets can lead to specific therapeutic courses) or even similarity between different diseases (treatments that work for one disease might also work for a similar one). Also, the natural extension of current models in medicine could lead to the development of computational reasoning tools specific to medicine with advanced cognitive computing capabilities and data sets as complete as possible. Such platforms could mine real-time clinical data, taking advantage of -omics, biomarkers, biomedical data and others, providing real-time patient services.
Tudor Oprea discussed how artificial intelligence and machine learning are revolutionizing drug discovery and presented his company's suite of LLM models, orchestrated by an artificial intelligence "super agent" that analyzes over 280 million articles, clinical trials and patents to find new treatments, faster. He detailed how this system works and presented, based on case studies, the advantages of its use, even offering access to all those interested in the DRUGINTELLM expert system through the OncoGen Center website.
For a short period of time, Prof. Tudor Oprea made that software available to the public at this link.