Mathematical AI for drug discovery

Abstract: Artificial intelligence (AI) has fundamentally changed the landscape of science, technology, industry, and social media in the past few years. It holds a great future for discovering new drugs significantly faster and cheaper. However, AI-based drug discovery encounters obstacles arising from the structural complexity of protein-drug interactions and the high dimensionality of drug candidates’ chemical space. We tackle these challenges mathematically. Our work focuses on reducing the biomolecular complexity and dimensionality in AI. We have introduced evolutionary de Rham-Hodge, algebraic topology, and persistent spectral graph theory to obtain high-level abstractions of protein-drug interactions and thus significantly enhance AI's ability to handle excessively large datasets of complex biomolecules in drug discovery. Using our mathematical AI approach, my team has been a top winner in D3R Grand Challenges, a worldwide annual competition series in computer-aided drug design and discovery in the past three years. I will briefly discuss Math and AI-based drug repositioning for COVID-19.

general mathematics

Audience: researchers in the topic

( slides )

Comments: the meeting ended

---

Global Seminar on Mathematical Modeling and Applications

Series comments: Description: Occasional seminar on mathematical modeling and related topics

In light of the worldwide cancellations of meetings workshops and conferences many of us are forced to switch to online mode of communication. Thus following the example of many other initiatives we decided to organize a Global Seminar on Mathematical Modeling and Applications.

Zoom links will appear on the seminar website 15 minutes before talks begin so that we may convene a little earlier, get acquainted and/or catch-up.

Additionally here you can subscribe to the mailing list.

---