Our work addresses the problem of predicting the highest priority functional group present in an organic molecule. Functional Groups are groups of bound atoms that determine the physical and chemical properties of organic molecules. In the presence of multiple functional groups, the dominant functional group determines the compound's properties. Fourier-transform Infrared spectroscopy (FTIR) is a commonly used spectroscopic method for identifying the presence or absence of functional groups within a compound. We propose the use of a Deep Convolutional Neural Networks (CNN) to predict the highest priority functional group from the Fourier-transform infrared spectrum (FTIR) of the organic molecule. We have compared our model with other previously applied Machine Learning (ML) method Support Vector Machine (SVM) and reasoned why CNN outperforms it.
Figures & tables
Functional Group
Number of Data samples
Carboxylic
537
Ester
333
Amide
948
Nitrile
371
Aldehyde
357
Ketone
183
Table 1: The class frequencies of highest priority functional groups present in our dataset.
No.
1600 (A)
3200 (B)
4730 (C)
SVM
CNN
SVM
CNN
SVM
CNN
1
64.51
68.85
66.18
75.01
69.12
79.70
2
63.31
69.53
66.54
74.20
68.55
77.77
3
64.40
67.32
65.93
74.51
68.30
76.89
4
63.56
70.49
67.45
73.89
70.63
79.84
5
63.87
68.14
67.62
73.97
78.80
68.68
Table 2: Accuracy of the ML models on (A) 1600 molecules randomly sampled from our dataset, (B) 3200 molecules randomly sampled from our dataset and (C) entire dataset of 4730 molecules.
K
SVM
CNN
1
68.42
78.43
2
80.1
88.20
3
89.2
94.3
Table 3: Accuracy for Top K predictions made by ML models on the dataset, that is the percentage of predictions in which the actual highest priority functional group is among the top k predictions (based on probability) made by the ML model.
No.
SVM
CNN
1
61.38
72.20
2
61.45
71.59
3
65.08
73.09
4
61.01
71.83
5
61.59
71.38
6
62.05
71.46
Table 4: Accuracy of ML models on undersampled data, i.e, randomly sample 200 examples of each class and form a dataset on which we perform Stratified 10 fold test.
SVM
RF
CNN
85.96
86.02
90.32
Table 5: Accuracy on applying ML models on previous paper’s [ 21 ] dataset.
Training Data
SVM
CNN
DATA-NEW
68%
74%
DATA-OLD
50%
59%
Table 6: Comparing accuracy of ML models, trained on different data.
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.
Yusen Tan, Hongyu Zhan, Hai-tao Yu +3
The Hong Kong University of Science and Technology (Guangzhou) · Westlake University · University of Science and Technology of China +1
Automated molecular structure elucidation from infrared (IR) spectroscopy data has seen significant advancements in recent years, but its broad applicability is limited by a reliance on pre-determined chemical formulas provided as auxiliary model inputs. This limits model predictions to isomer identification rather than full molecular structure prediction. Although transformer models have been shown to identify molecular isomers with high accuracy, their reliability for unconstrained structure elucidation is comparatively low and poorly understood. In this work, we propose and evaluate key modifications to the traditional encoder-decoder transformer. To better address the vast chemical space of the unconstrained problem, we implement a novel Mixture-of-Experts (MoE) decoder module that utilizes non-additive aggregation via linear-order statistics and the Choquet integral. We further modify the transformer to utilize these non-additive operators when aggregating spectral representations as well. Together with an auxiliary contrastive alignment loss term, these enhancements improve Top-K prediction accuracy by over 10 percentage points compared to baseline IR-only models. Through sub-structure fragment analysis of molecular predictions, we further confirm that infrared spectra encode the vast majority of relevant chemical information, implying that the higher performance of isomer-ranking models is largely due to underrepresented or overlapping absorption bands for molecules in the explored chemical space. Ultimately, by demonstrating the efficacy of automated molecular structure elucidation from measured IR spectra, this work serves to significantly broaden the utility of AI in analytical chemistry.
Ethan J. Mick, Campbell A. Sweet, Matthias J. Young +1
Electrical Engineering and Computer Science University of Missouri Columbia, MO 65211 · Chemical and Biomedical Engineering University of Missouri Columbia, MO 65211
Multi-RF Fusion achieves a test ROC-AUC of 0.8476 +/- 0.0002 on ogbg-molhiv (10 seeds), placing #1 on the OGB leaderboard ahead of HyperFusion (0.8475 +/- 0.0003). The core of the method is a rank-averaged ensemble of 12 Random Forest models trained on concatenated molecular fingerprints (FCFP, ECFP, MACCS, atom pairs -- 4,263 dimensions total), blended with deep-ensembled GNN predictions at 12% weight. Two findings drive the result: (1) setting max_features to 0.20 instead of the default sqrt(d) gives a +0.008 AUC gain on this scaffold split, and (2) averaging GNN predictions across 10 seeds before blending with the RF eliminates GNN seed variance entirely, dropping the final standard deviation from 0.0008 to 0.0002. No external data or pre-training is used.