Explainable AI in Drug Discovery: Why "Why" Matters
A model that predicts activity is useful. A model that tells a chemist which substructure drives that activity is something you can act on.
2025
Anatomy of an AI-Driven Molecular Design Workflow
How generative models, scoring and chemical filters come together into a practical loop that turns a target into testable molecules.
2020
Highlights of the Month: May 2020
Keywords: ICLR 2020; multitask learning with fastai; fastai + huggingface; learn2learn
A walkthrough of TMAP
The notebook version is here.
Highlights of the Month: April 2020
Key words: JupyterLab; NMR; WSL; NLP tutorial/recipes
My RDKit Cheatsheet
Cheatsheet for RDKit package in python: (1) Draw molecules in jupyter enviroment; (2) use with Pandas Dataframe (3) Descriptors/Fingerpri...
Highlights of the Month: March 2020
Key words Sub-Word Tokenization; Ligand-Based Virtual Screening; Meta Learning; ELECTRA;
Highlights of the Month: February 2020
Key words for this month: Reformer; TMAP; libmolgrid;
Highlights of the Month: January 2020
Key words for this month: Functional Groups; Unsupervised/Self-supervised representation leaning; Generative Models;
2019
DGL Walkthrough 01: Data
A walk through of DGL library. Deep Graph library (DGL) is a Python package built for easy implementation of graph neural network model f...
Create an Interactive Visualization for Approved Small Molecule Drugs
In the post, I want to generate an interactive visualization of a chemical space. Each point in the map represents a molecule and close p...