About me

My name is Benoit Baillif, I am currently a PhD student in the Bender group from the department of Chemistry at the University of Cambridge. My main research interest is to leverage three-dimensional (3D) molecular structure data to enhance drug discovery. In my first project, I used atomistic neural networks to bias conformer ensemble towards bioactive-like conformations. In my current project, I will benchmark existing 3D deep molecular generation methods, both unconditioned generation (diverse conformations for diverse molecules) and conditioned generation (e.g. pocket conditioning to generate active molecule directly in target-bound poses). I also worked on deep conformation generation in my first year, and worked on a side project exploring the use of the Smooth Overlap of Atomic Positions (SOAP) descriptor for protein-ligand prediction. Besides all this interesting work, I am generally passionate about new technologies, science fiction, board games and recently salsa dancing.

My Projects

GitHub repositories that I've built.

My Interests

Topics that I want to learn more about.

Python

#

Machine learning

#

Deep learning