Generation of 3D geometries and graphs aided by efficient spectral methods

The project aims to facilitate the research and development of new drugs by employing machine learning methods to generate new molecules. This includes understanding various properties about known molecules by training deep learning models for the purpose of molecular generation. The gained understanding of molecules will be used to improve existing models and generate novel molecules with high likelihood of satisfying given properties. Valence Discovery will benefit from the better molecular generation models since they can help reduce the cost and time involved with new drug-discovery or drug optimization objectives. Meaning that they can improve the quality of the drugs, reduce the cost of development and the side effects. This research will utilize expertise from graph representation learning, 3D geometries, spectral learning, generative modeling, and computational drug discovery.

Faculty Supervisor:

Reihaneh Rabbany

Student:

Partner:

Valence Discovery Inc

Discipline:

Computer science

Sector:

Professional, scientific and technical services

University:

McGill University

Program: