Skip to content

Bridging the Computational-Experimental Domain Gap: An Autoencoder-Based Machine Learning Workflow for Raman Spectra Characterization of Amino Acid Mixtures.

Aug 2026 · Journal of Physical Chemistry A · Vol 130 33, pp. 6633-6645 · 0 citations · 57 references
Medicine

TL;DR

An autoencoder-based workflow is developed that enables machine-learning classifiers trained exclusively on computational Raman spectra to identify the dominant amino acid component in experimental mixtures and demonstrates that baseline correction combined with deterministic autoencoder-based domain transformation can effectively bridge computational and experimental Raman spectra.

Abstract

Raman spectroscopy offers rapid, nondestructive molecular characterization, but machine-learning analysis of experimental Raman spectra is often limited by the cost and difficulty of constructing sufficiently large, reproducible experimental data sets. Theoretical spectra generated by density functional theory provide an attractive alternative; however, discrepancies in peak positions, intensities, line shapes, and background signals create a substantial domain gap between computed and measured spectra. This study develops an autoencoder-based workflow that enables machine-learning classifiers trained exclusively on computational Raman spectra to identify the dominant amino acid component in experimental mixtures. Raman spectra of the 20 common proteinogenic amino acids were calculated using density functional theory over the 400-1800 cm1 region. Binary mixtures were generated by randomly combining two amino acids at variable ratios, producing 20,000 computational spectra for training Random Forest and XGBoost classifiers. Both classifiers achieved approximately 99.5% accuracy on held-out computational spectra, confirming that the dominant component could be identified from theoretical spectral features. Nevertheless, direct application to experimental phenylalanine-glutamate mixtures resulted in essentially 0% accuracy, demonstrating that baseline correction alone was insufficient to overcome the computational-experimental discrepancy. To address this limitation, experimental spectra were first processed using asymmetrically reweighted penalized least-squares baseline correction and subsequently transformed toward their corresponding computational representations using variational autoencoder, standard autoencoder, and Gaussian-mixture variational autoencoder architectures. Combining baseline correction with the variational autoencoder constrained predictions to the two amino acids present in the mixtures but yielded only 50% accuracy. The Gaussian-mixture variational autoencoder produced no meaningful further improvement. In contrast, removal of latent-space Gaussian regularization allowed the standard autoencoder to minimize the direct reconstruction discrepancy between experimental inputs and computational targets, increasing classification accuracy to 99.645% for Random Forest and 99.395% for XGBoost. These results demonstrate that baseline correction combined with deterministic autoencoder-based domain transformation can effectively bridge computational and experimental Raman spectra. The proposed workflow substantially reduces dependence on large experimental training data sets and provides a practical framework for low-cost Raman-based characterization of amino acid mixtures, with potential applications in biomedical analysis, clinical screening, and chemically informed spectroscopic machine learning.

View source

Similar papers

Open access Oct 2026

Experimental FTIR and Raman Vibrational Analysis of 4-Amino-2-Chlorobenzonitrile: Cross-Technique Comparison and a Framework for Computational and Machine-Learning Validation

This study organizes the recorded Fourier transform infrared and laser Raman bands of 4-amino-2chlorobenzonitrile into an experimental benchmark for computational vibrational analysis and interpretable machine learning. The available measurements comprise peak assignments obtained in potassium bromide, Nujol and Raman...

Surabhi Singhal · 0 citations
Open access Sep 2026

Learning from tandem mass spectra at scale with a self-supervised foundation model for proteomics

InstaNovo-FM is presented, a self-supervised foundation model for bottom-up proteomics trained to reconstruct masked regions of tandem mass spectra and establishes a unified representation space for peptide fragmentation spectra, enabling robust transferability across the proteomics ecosystem.

M. Nieuwoudt, Marco Reverenna, Divanisha Patel et al. · 1 citation · ⚡1
Open access Sep 2026

A machine-learning-driven dataset of 144,111 polycyclic aromatic hydrocarbon infrared spectra for interpreting aromatic infrared bands

The interpretation of interstellar aromatic infrared bands (AIBs) relies on reference spectral libraries, but the chemical space covered by existing databases remains incomplete relative to the vast structural diversity of polycyclic aromatic hydrocarbons (PAHs) expected in the interstellar medium. We present a large-s...

Xing-Hong Mai, Zu-Jia Lu, Zhao Wang · 0 citations
#machine learning Preprint Oct 2026

Wavenumber-Aware Deep Learning for Robust Analysis of Raman Spectra

In controlled comparisons, wavenumber-based encoding outperformed index-based and learned encodings under partial observation and retained its performance on sampling grids never seen during training, where index-based encoding degraded.

Jihwan Kim, Jaeseong Heo, Yujin Kim et al. · 0 citations
Open access Oct 2026

MARS: Machine-Learned Force Field Framework for Automated Conformational Sampling, Vibrational Spectroscopy, and Microsolvation.

The results establish MARS as a transferable and accessible platform for ensemble-based molecular modeling with machine-learned potentials as a transferable and accessible platform for ensemble-based molecular modeling with machine-learned potentials.

Sergio Suárez-Dou, Miguel Gallegos, F. Brunig et al. · 0 citations
Oct 2026

The Delta-ML IR Spectroscopist: Structured Delta-Learning between Harmonic DFT Infrared Spectra

Accurate theoretical infrared spectra support molecular identification and high-throughput screening, but higher-level harmonic density functional theory (DFT) calculations are costly at scale; we therefore test whether a low-cost PBEh-3c harmonic spectrum can serve as a physical prior for predicting an ωB97X-D3/def2...

Rhys McAlister, L. McKemmish · 0 citations

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.