When machine learning models learn chemistry I: quantifying explainability with matched molecular pairs

2026/01/01 by Kerrin Janssen, Jan M. Wollschläger, Jonny Proppe +1 · 1 voice
Computer Science · Materials Science · #Computational Drug Discovery Methods #Explainable Artificial Intelligence (XAI) #Machine Learning in Materials Science

paper · pdf · doi:10.1039/d5dd00398a

openalex publication_date 2026/01/01 · openalex created_date 2026/02/10 · openalex updated_date 2026/05/21

Abstract

Explainability methods in machine learning-driven research are increasingly being used, but it remains challenging to assess their reliability without deeply investigating the specific problem at hand.

Citations

• Drug discovery with explainable artificial intelligence
• Trust in Automation: Designing for Appropriate Reliance
• QSAR without borders

Discussions

• What happens when #machinelearning models learn chemistry? Find out in this two part bonus👇 Part I introduces #Python-based Workflow for Interpretability Scoring using matched molecular Pairs: WISP👉 [bsky, 1 points, 0 comments]

Related

• Explainability Techniques for Chemical Language Models
• MEG: Generating Molecular Counterfactual Explanations for Deep Graph Networks
• MoleculeNet: A Benchmark for Molecular Machine Learning
• Explaining Deep Graph Networks with Molecular Counterfactuals
• Enhancing Chemical Explainability Through Counterfactual Masking