Towards Reasonable Molecular Structure Elucidation from Infrared Spectroscopy with Chemical Feedback
2026-08-17 • Machine Learning
Machine Learning
AI summaryⓘ
The authors found that current machine learning models often suggest wrong molecular structures when interpreting infrared (IR) spectra, especially when the predicted structures don't match the known molecular formula or IR data. They created a new method called FIRMPO that uses chemical rules to help the models pick structures that better fit both the formula and IR signals. This method can work with many existing models and helps improve the accuracy of predicting the correct molecular structure from IR spectra. Their tests on popular IR datasets showed that FIRMPO performs better than previous approaches.
Infrared (IR) spectroscopyMolecular structure elucidationMachine learningMolecular formulaSpectral consistencyPreference optimizationChemical feedbackModel-agnosticStructure predictionIR datasets
Authors
Yusen Tan, Hongyu Zhan, Hai-tao Yu, Changxi Chi, Wenjie Du, Jun Xia
Abstract
Infrared (IR) spectra provide characteristic signals of molecular structure, which are often interpreted by experts via functional-group identification or library matching, making the process time-consuming and ambiguous. Recent machine learning methods have made progress in molecular structure elucidation using molecular formulas and IR spectra. However, these models often infer unreasonable candidate molecular structures, including top-ranked predictions. More specifically, the molecular formula implied by a candidate structure often fails to match the input molecular formula, and the candidate's theoretical IR spectrum is often inconsistent with the observed IR spectrum. To address these issues, we propose Formula- and IR-Matched Preference Optimization (FIRMPO), a general and plug-and-play chemical feedback-driven preference optimization framework for molecular structure elucidation. FIRMPO incorporates chemical feedback as preference signals based on exact molecular formula matching and IR spectral consistency to guide reasonable structure predictions. Unlike generic preference optimization methods, FIRMPO is tailored to molecular structure elucidation while remaining model-agnostic, enabling it to be readily integrated with different structure prediction models in this class. This encourages models to prioritize structures that satisfy the chemical feedback, leading to a substantial improvement in the accuracy of top-ranked predictions. Extensive experiments on three widely used IR datasets show that FIRMPO significantly improves molecular structure elucidation accuracy over existing baselines.