De Novo Structure Prediction from Tandem Mass Spectra: Algorithms, Benchmarks, and Limitations
Molecules · 25.02.2026 · 31 (5) · 769
DOI: 10.3390/molecules31050769 ↗Kurz gefasst
Übersichtsarbeit in der Zeitschrift Molecules, online erschienen am 25.02.2026. Verfasst von Mark Yu Schneider und 6 weiteren Autor:innen. Beteiligte Einrichtung: Research Center of the Artificial Intelligence Institute, Innopolis University, 420500 Innopolis, Russia. Die Arbeit wurde bisher 4-mal zitiert (Crossref) und nennt 114 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: The study was supported by the Ministry of Economic Development of the Russian Federation, Ministry of Economic Development of the Russian Federation.
Abstract (Originalsprache)
The identification of unknown molecules from analytical data remains a fundamental challenge in chemistry, with critical implications for drug discovery, metabolomics, and natural product research. While tandem mass spectrometry provides rich structural fingerprints, most spectra are absent from reference libraries, spurring the development of de novo generative models. However, their true accuracy has been difficult to assess. Our critical analysis reveals that state-of-the-art models achieve only 4.1% top-10 accuracy on rigorously leakage-controlled benchmarks like MassSpecGym. This sobering figure stands in stark contrast to earlier, overly optimistic reports, a discrepancy we attribute to pervasive data leakage in naive data splits. This review traces the field’s rapid evolution through three architectural eras: from fingerprint-conditioned RNN pipelines to end-to-end sequence models and, most recently, to graph-native diffusion under molecular-formula constraints. We demonstrate that explicitly conditioning generative models on a molecular formula significantly improves exact-match accuracy compared to unconstrained baselines. Crucially, our analysis distinguishes between two experimentally relevant paradigms: formula-conditioned generation for true unknown discovery and scaffold-based generation for hypothesis-driven research. While the latter shows high potential with oracle scaffolds, its performance drastically drops with predicted ones, revealing a critical bottleneck. To build the next generation of reliable tools, we propose a clear roadmap centered on standardized, leakage-aware benchmarking and transparent reporting.
Schlagwörter
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Research Center of the Artificial Intelligence Institute, Innopolis University, 420500 Innopolis, Russia.
ORCID-Profile: Mark Yu Schneider, Daniil D Kholmanskikh, Kirill Ya Romanov, Elena A Perekina, Sergei A Nikolenko, Ruslan Yu Lukin, Ivan V Golov
Förderung
- The study was supported by the Ministry of Economic Development of the Russian Federation
- Ministry of Economic Development of the Russian Federation
Zitieren
APA
Schneider, M. Y., Kholmanskikh, D. D., Romanov, K. Y., Perekina, E. A., Nikolenko, S. A., Lukin, R. Y., & Golov, I. V. (2026). De Novo Structure Prediction from Tandem Mass Spectra: Algorithms, Benchmarks, and Limitations. Molecules, 31(5), 769. https://doi.org/10.3390/molecules31050769
Vancouver
Schneider MY, Kholmanskikh DD, Romanov KY, Perekina EA, Nikolenko SA, Lukin RY, et al. De Novo Structure Prediction from Tandem Mass Spectra: Algorithms, Benchmarks, and Limitations. Molecules. 2026;31(5):769. doi:10.3390/molecules31050769
BibTeX
@article{Schneider2026,
author = {Schneider, Mark Yu and Kholmanskikh, Daniil D and Romanov, Kirill Ya and Perekina, Elena A and Nikolenko, Sergei A and Lukin, Ruslan Yu and Golov, Ivan V},
title = {De Novo Structure Prediction from Tandem Mass Spectra: Algorithms, Benchmarks, and Limitations},
journal = {Molecules},
year = {2026},
volume = {31},
number = {5},
pages = {769},
doi = {10.3390/molecules31050769}
}
EUVIMED