Graph Representation Learning for Drug Discovery

Traditionally, drug discovery is a long and costly process, which consists of many stages and on average takes a total of more than 10 years, because it screens huge amount of drug candidates by conducting chemical and biological experiments which are expensive and time consuming.
This project aims to apply graph representation learning technologies to drug discovery. These technologies suit the drug domain since molecules can be well represented as graph structures. They can be used to predict the properties of drug candidates, to predict how to synthesize drug molecules, and to predict how to generate new drug candidates. All based on AI models that learns from large amount of drug molecule data that exists in the pharmaceutical industry, without conducting chemical experiments. Thus, it benefits the partner organization both in terms of reduced cost and shortened time required for discovering new drugs.

Faculty Supervisor:

Di Niu

Student:

Partner:

Ark (Canada) Intelligence Ltd.

Discipline:

Computer science

Sector:

Professional, scientific and technical services

University:

University of Alberta

Program: