Skip to main content
Loading Events

« All Events

  • This event has passed.

Colloquium Series: Dr. Alex Trophsa, UNC

Tuesday, November 1, 2022 @ 4:00 pm - 5:00 pm

Join us at Chapman Hall 125 for our Colloquium Series. Dr. Alex Tropsha, distinguished professor at UNC will be joining us to present his research on Applications of Machine Learning and Artificial Intelligence approaches for rational discovery of chemicals and materials.
Research Synopsis:

One of the key objectives of this seminar is to share our research and explore collaborative opportunities with members of the APS department. In general, our current projects build in two clear trends that have emerged concurrently and synergistically in the last decade in multiple disciplines including chemistry and materials science: data science, fueled by rapid accumulation of research data stored in multiple specialized databases and the development of novel machine learning (ML) and artificial Intelligence (AI) algorithms capable of analyzing and modeling this big data. I will reflect on recent methodological and applied research in our group in three connected areas: knowledge discovery in biomedical databases; discovery of novel bioactive compounds; and discovery of novel materials. I will introduce the concepts of biomedical ‘graph knowledgebases’ and knowledge graph mining that support our novel question-answering system termed Reasoning Over Biomedical Objects linked in Knowledge-Oriented Pathways (ROBOKOP), and I will describe the use of ROBOKOP for drug repurposing studies. I will also discuss our studies on the development of a deep learning/reinforcement learning methodology for de novo drug discovery and present recent results on using this approach for structure based discovery of novel broad spectrum antiviral compounds. Finally, I will talk about our use of Quantitative Structure-Property Relationship (QSPR) modeling approaches for designing novel materials for solar energy applications. In conclusion, I will summarize the importance of continuous data cycle integrating data-generating and data modeling research illustrated by the emergence of automated chemistry labs.


Tuesday, November 1, 2022
4:00 pm - 5:00 pm
Event Category:
Comments are closed.