Research

Research Interests

Research Agenda

Grants

Deep generative and predictive models in drug discovery, 2020

In 2020, I was awarded with a grant by Whitireia and Weltec Research and Innovation Fund entitled "Deep generative and predictive models in drug discovery". In the project, I am the sole principal investigator.

The project developed novel methods for lead discovery by developing (1) a novelty-oriented generative model using deep learning, (2) a series of predictive models for druglikeness-based virtual screening, and (3) a GUI friendly web-based system and mobile app supporting drug discovery.

The project has been successfully completed with one high impact-factor journal paper published at the Journal of Chemical Information and Modeling (IF: 4.956) and one journal paper which is under review at Proteomic (IF: 3.984).

The web-based system is now maintained and running at website. In this system, the users can choose to perform molecular generation with three options including random generation, property-based generation and scaffold-based generation.

For random generation, with each generation batch, to-be-produced molecules are randomly sampled from the chemical space. For property-based generation, with each generation batch, to-be-produced molecules are randomly sampled from the chemical space but the output molecules will be driven to meet desired properties. For scaffold-based generation, with each generation batch, to-be-produced molecules will carry novel substructures (randomly sampled from the chemical space) and partial substructures (inherited from input molecules).

AI framework to predict drug-target interaction, 2022

Publications

My full list of publications can be found here.