Machine-learning-enhanced Optical Photothermal infrared (O-PTIR) spectroscopy for rapid simultaneous identification of multiple metabolites to monitor biotransformations: proof of concept on Brewer's spent grain (BSG)
Abstract
Background Monitoring fermentation in chemically complex matrices remains analytically challenging because multiple metabolites evolve simultaneously. Conventional assays are targeted and sequential, and only partial annotation is often available in real samples. In this study, we present a masked multi-label analytical workflow combining optical photothermal infrared spectroscopy (O-PTIR) with a one-dimensional convolutional neural network (1D-CNN) for rapid screening of brewer's spent grain fermentations. The methodology couples a library of standard compound spectra with transfer learning on partially labelled fermentation spectra, while masked optimisation avoids penalising unannotated classes during fine-tuning. Results A library of ten reference compounds and their mixtures was first used to train the model, which was then adapted to fermented samples using qualitative labels for starch, glucose, lactic acid and kojic acid. Supernatants from non-inoculated, separated and sequential bacterial/fungal fermentations were analysed after minimal sample preparation. The pretrained model showed excellent internal validation on the standard library. The fine-tuned 1D-CNN retained high discriminative performance (macro-F1 = 0,950). Benchmarking against one-vs-rest PLS-DA and SVM models using the same held-out BSG validation split showed that the nonlinear models outperformed the linear chemometric baseline. Predicted temporal profiles were consistent with assay-derived phase transitions, notably glucose depletion and lactic acid accumulation during the bacterial stage, followed by starch disappearance and kojic acid appearance during the fungal stage. Significance These results support the use of multi-label O-PTIR workflow to provide rapid and label-free identification of key metabolic phases in partially labelled, heterogeneous matrices, and highlight its potential as a transferable analytical strategy for bioprocess monitoring.
Related
- Label-free identification and classification of leukocyte subpopulations in peripheral blood smears by optical photothermal infrared spectroscopy
- Monitoring Early Infection Dynamics in Live Spodoptera frugiperda (Sf) 9 Cells via Optical Photothermal Infrared (OPTIR) Spectroscopy.
- Evaluation of filter substrates for concurrent optical photothermal infrared and Raman spectroscopy in microplastic analysis
- Rapid and simultaneous identification and quantification of seven commonly consumed meat species in food products using two-tube TaqMan probe-based multiplex real-time PCR assays
- Simultaneous CEA and CDH-17 Detection from Cancer Extracellular Vesicles Via Machine Learning Enabled Bio Silent Azide-Infrared Spectroscopy
BibTeX
@article{lagneaux2026machinelearningenhanced,
title = {{Machine-learning-enhanced Optical Photothermal infrared (O-PTIR) spectroscopy for rapid simultaneous identification of multiple metabolites to monitor biotransformations: proof of concept on Brewer's spent grain (BSG)}},
author = {Paul Lagneaux and Muhammad Hassan and Pascale Winckler and Yves Waché},
journal = {Analytica Chimica Acta},
year = {2026},
month = aug,
doi = {10.1016/j.aca.2026.346159},
url = {https://institut-agro-dijon.hal.science/hal-05749000v1},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Machine-learning-enhanced Optical Photothermal infrared (O-PTIR) spectroscopy for rapid simultaneous identification of multiple metabolites to monitor biotransformations: proof of concept on Brewer's spent grain (BSG)
AU - Paul Lagneaux
AU - Muhammad Hassan
AU - Pascale Winckler
AU - Yves Waché
PY - 2026
DA - 2026/08/22
JO - Analytica Chimica Acta
DO - 10.1016/j.aca.2026.346159
UR - https://institut-agro-dijon.hal.science/hal-05749000v1
AB - Background Monitoring fermentation in chemically complex matrices remains analytically challenging because multiple metabolites evolve simultaneously. Conventional assays are targeted and sequential, and only partial annotation is often available in real samples. In this study, we present a masked multi-label analytical workflow combining optical photothermal infrared spectroscopy (O-PTIR) with a one-dimensional convolutional neural network (1D-CNN) for rapid screening of brewer's spent grain fermentations. The methodology couples a library of standard compound spectra with transfer learning on partially labelled fermentation spectra, while masked optimisation avoids penalising unannotated classes during fine-tuning. Results A library of ten reference compounds and their mixtures was first used to train the model, which was then adapted to fermented samples using qualitative labels for starch, glucose, lactic acid and kojic acid. Supernatants from non-inoculated, separated and sequential bacterial/fungal fermentations were analysed after minimal sample preparation. The pretrained model showed excellent internal validation on the standard library. The fine-tuned 1D-CNN retained high discriminative performance (macro-F1 = 0,950). Benchmarking against one-vs-rest PLS-DA and SVM models using the same held-out BSG validation split showed that the nonlinear models outperformed the linear chemometric baseline. Predicted temporal profiles were consistent with assay-derived phase transitions, notably glucose depletion and lactic acid accumulation during the bacterial stage, followed by starch disappearance and kojic acid appearance during the fungal stage. Significance These results support the use of multi-label O-PTIR workflow to provide rapid and label-free identification of key metabolic phases in partially labelled, heterogeneous matrices, and highlight its potential as a transferable analytical strategy for bioprocess monitoring.
SN - 0003-2670
ER -