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

Kerrin Janssen; Jan M. Wollschläger; Jonny Proppe; Andreas H. Göller

doi:10.1039/d5dd00398a

January 1, 2026Digital DiscoveryOpen Access

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

View Full Paper

Ask AI

Bookmark

Share

Key Points

The aim is to investigate the explainability of machine learning methods in the context of chemistry, focusing on matched molecular pairs.
Utilized machine learning algorithms to analyze molecular pairs.
Assessed different explainability methods to evaluate their reliability.
Conducted comparative analysis to identify effective strategies for explainability.
Identified challenges in assessing the reliability of explainability methods.
Demonstrated variations in effectiveness of explainability techniques across different molecular pairs.
Reported a need for deeper investigation into specific chemistry problems.

Implication

Machine learning models quantify explainability of molecular pairs, suggesting enhanced evaluation methods are needed.

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.

Discussion

No takes yet. Share an insight, caveat, or question.

Cite This Study

Janssen et al. (2026) studied this question.

synapsesocial.com/papers/698c1cd3267fb587c655f99c https://doi.org/10.1039/d5dd00398a

View Full Paper

Ask AI

Bookmark

Share

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

Key Points

Implication

Abstract

Discussion

Cite This Study

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

Key Points

Implication

Abstract

Discussion

Cite This Study

Keep reading

Keep reading