Dries Van Rompaey
dries-vr.bsky.social
Dries Van Rompaey
@dries-vr.bsky.social
Machine learning for molecular discovery @JNJ
New blog post is live! Join me in taking a look at the intuition behind the tried and tested Matched Molecular Series method, coding up an implementation leveraging polars, and exploring some matched series!

driesvr.github.io/2025/01/27/m...
Matching Molecular Series with MatchMolSeries.py
“What else can we put on that position to improve potency?” If you’ve worked in small molecule discovery, you’ll probably have heard some variation of that question before. Replacing R-groups tends to...
driesvr.github.io
January 28, 2025 at 7:05 PM
Free-Wilson Analysis is an awesome tool for both compound generation and data analysis. In this very first blog post, I try to remedy some of the edge cases that can pop up: driesvr.github.io/2025/01/06/f...
Free-Wilson edge-cases
I’ve always been a fan of Free-Wilson Analysis. It’s just such a useful tool: it’s easy-to-use, it helps you understand SAR better and it generates all potentially interesting combinations of your com...
driesvr.github.io
January 9, 2025 at 8:46 PM
Can we already start thinking about dose at the point of design? We discuss how we try to tackle this challenge and take a closer look at the impact of the accuracy of individual predictions.

pubs.acs.org/doi/abs/10.1...
Toward Dose Prediction at Point of Design
Human dose prediction (HDP) is a useful tool for compound optimization in preclinical drug discovery. We describe here our exclusively in silico HDP strategy to triage compound designs for synthesis a...
pubs.acs.org
December 12, 2024 at 8:01 PM
Reposted by Dries Van Rompaey
Here's a Cheminformatics Starter
pack! Let me know if you would like to be added! And more importantly, do encourage people to join the blue skies :D
November 17, 2024 at 4:43 AM