Cycle-MS: A Closed-Loop End-to-End Framework for Mass Spectrometry Structure Elucidation.

Area of Science:

• Computational chemistry
• Analytical chemistry
• Structural biology

Background:

• Mass spectrometry (MS) is crucial for molecular analysis.
• Elucidating complex molecular structures from MS spectra is challenging.

Purpose of the Study:

• Introduce Cycle-MS, an end-to-end framework for MS-based structure elucidation.
• Improve accuracy and interpretability in predicting molecular structures from spectra.

Main Methods:

• Unify inverse spectrum-to-molecule prediction with forward spectrum reconstruction.
• Utilize cycle-consistency losses to refine latent representations.
• Conduct interpretability analysis to assess feature prioritization.

Main Results:

• Cycle-MS demonstrates enhanced accuracy in predicting atomic compositions, molecular formulas, scaffolds, and functional groups.
• Achieve robust recovery of molecular scale and key chemical features.
• Prioritize structurally diagnostic fragments over nonspecific abundant peaks for chemical plausibility.

Conclusions:

• Cycle-MS provides a robust, accurate, and interpretable framework for MS data.
• Enable fully automated, high-precision MS-based structure elucidation.
• Advance the field of computational molecular structure determination.