MassSpecGym in the Wild: Uncovering and Correcting Evaluation Pitfalls in AI-Driven Molecule Discovery

2026-06-17

AI Summary

This study addresses widespread reliability issues in the evaluation of AI-driven molecular discovery from mass spectrometry data, where practices such as data leakage, shortcut learning, and implementation errors frequently lead to inflated performance claims. The authors conduct the first systematic audit of 26 papers benchmarked on MassSpecGym, identifying and categorizing three distinct failure modes in model evaluation. Through rigorous reproduction experiments, they quantify the impact of these flaws on reported metrics. To remedy these shortcomings, the work introduces a standardized evaluation protocol, releases an updated benchmark—MassSpecGym v1.5—and provides an open-source toolkit. Their analysis reveals that at least 17 of the reviewed studies exhibit significant evaluation defects, establishing a more robust and reproducible foundation for future AI models in mass spectrometry-based molecular discovery.

Abstract

Reliable benchmarking is critical for developing machine learning models for tandem mass spectrometry (MS/MS) based molecule discovery. Subtle issues in experimental design and model evaluation procedures can degrade the trustworthiness of such benchmarks and lead to erroneous conclusions. We conduct a thorough review of model evaluation issues in the recent MS/MS machine learning literature, using the standard MassSpecGym benchmark suite as a case study to illustrate the impact of these issues. We find evaluation issues in at least 17 of 26 papers reporting MassSpecGym benchmark results in the first year of its adoption. We isolate three classes of failures: (i) data leakage, (ii) shortcut learning, and (iii) implementation bugs and metric divergence. Through extensive experimentation and code replication, we quantify the impact of these issues and show how they corrupt the evaluation standards MassSpecGym was designed to enforce. We distill our findings into recommendations generalizable to MS/MS challenges, benchmarks, and custom evaluation setups. We also release MassSpecGym v1.5, an implementation of our recommendations in the MassSpecGym benchmarking suite which addresses the failure modes identified in this audit. MassSpecGym v1.5 is publicly available at https://github.com/pluskal-lab/MassSpecGym.

Problem

• mass spectrometry
• machine learning
• benchmarking
• evaluation pitfalls
• molecule discovery

Innovation

• evaluation pitfalls
• data leakage
• shortcut learning
• mass spectrometry
• benchmarking