analiza aiqbench
MSAlign advances metabolite identification by aligning frozen foundation models for mass spectra and molecular structures, offering a lightweight solution with state-of-the-art performance. The work also critically addresses evaluation methodology by quantifying distribution shifts to mitigate data leakage in existing benchmarks. This research provides a robust framework for reproducible machine learning applications within computational chemistry and drug discovery.
MSAlign: Aligning Molecule and Mass Spectra representations for Metabolite Identification
arXiv:2605.19752v2 Announce Type: replace Abstract: Accurately identifying metabolites i.e. small molecules from mass spectrometry data remains a core challenge in metabolomics, with broad applications in drug discovery, environmental analysis, and clinical research. We address the Molecule Retrieval task, which consists in recovering the chemical structure of a metabolite from its MS/MS spectrum given a set of candidate molecules. We make three contributions. First, we propose a unified framework encompassing recent approaches based on representation alignment and contrastive learning. Second, we introduce MSAlign, a lightweight model that achieves state-of-the art performances by aligning two frozen foundation models (DreaMS for mass spectra and MolDeBERTa for molecules) and demonstrate that a score fusion strategy further improves the performance for a very small computational cost. Third, we investigate a long-standing evaluation problem: data splitting strategies in molecule retrieval implicitly trade off data leakage against domain shift. We formalize this tension by introducing a quantitative measure of distribution shift, and use it to evaluate splitting strategies in existing benchmarks. All datasets, splits, candidate sets, and a unified implementation of MSAlign and baselines are publicly released to support reproducible research.
Abstract:Accurately identifying metabolites i.e. small molecules from mass spectrometry data remains a core challenge in metabolomics, with broad applications in drug discovery, environmental analysis, and clinical research. We address the Molecule Retrieval task, which consists in recovering the chemical structure of a metabolite from its MS/MS spectrum given a set of candidate molecules. We make three contributions. First, we propose a unified framework encompassing recent approaches based on representation alignment and contrastive learning. Second, we introduce MSAlign, a lightweight model that achieves state-of-the art performances by aligning two frozen foundation models (DreaMS for mass spectra and MolDeBERTa for molecules) and demonstrate that a score fusion strategy further improves the performance for a very small computational cost. Third, we investigate a long-standing evaluation problem: data splitting strategies in molecule retrieval implicitly trade off data leakage against domain shift. We formalize this tension by introducing a quantitative measure of distribution shift, and use it to evaluate splitting strategies in existing benchmarks. All datasets, splits, candidate sets, and a unified implementation of MSAlign and baselines are publicly released to support reproducible research.
Machine Learning (cs.LG)
arXiv:2605.19752 [cs.LG]
(or arXiv:2605.19752v2 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2605.19752 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 271 cuvinte
Subiecte conexe
Data Fusion and Contrastive Alignment for Unconstrained IR Molecular Structure Elucidation
arXiv · 30 iul. 2026
GyroNovo: Error-Guided Fragment Imputation with Mass-Aware Attention for \textit{De Novo} Peptide Sequencing
arXiv · 28 sept. 2026
MS-GPT: Rethinking MS/MS De Novo Structure Elucidation as Spectrum-Induced Posterior Querying of a Molecule-Language Model
arXiv · 28 iul. 2026
The Platonic Universe: Do Foundation Models See the Same Sky?
arXiv · 30 sept. 2026