On the use of TabPFN on mass spectrometry analysis of volatile organic compounds

Autores
Granitto, Pablo Miguel; Betta, Emanuela; Khomenko, Iuliia; Pedrotti, Michele; Romano, Andrea; Biasioli, Franco
Año de publicación
2025
Idioma
inglés
Tipo de recurso
artículo
Estado
versión publicada
Descripción
Volatile organic compounds (VOCs) are key markers in applications ranging from food quality assessment to medical diagnostics that can be profiled, for example, by gas chromatography–mass spectrometry (GC-MS) or by direct injection mass spectrometry (e.g. proton transfer reaction mass spectrometry). The common practice in both cases is to construct a tabular dataset from the raw measurements by performing peak extraction across samples and use statistical or machine learning methods to analyze it. However, modeling VOC profiles is particularly challenging due to high dimensionality, noise, and small sample sizes. In this study, we evaluate the Tabular Prior-data Fitted Network (TabPFN), a foundation model recently introduced for tabular data, across diverse VOC datasets. Without requiring task-specific training, TabPFN achieves state-of-the-art performance in both classification and regression tasks, outperforming classical machine learning methods for most datasets. We further explore new strategies to enhance TabPFN’s performance, including ensembling and fine-tuning, finding that a plain ensemble seems to be the best option in this setting. Our results demonstrate that TabPFN is a highly effective modeling tool for VOC profiles obtained with different analytical approaches. It offers robust predictions even in the data-scarce, high-variability scenarios typical of real-world workflows.
Fil: Granitto, Pablo Miguel. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Conicet - Rosario. Centro Internacional Franco Argentino de Ciencias de la Información y de Sistemas. Universidad Nacional de Rosario. Centro Internacional Franco Argentino de Ciencias de la Información y de Sistemas; Argentina
Fil: Betta, Emanuela. Instituto Agrario San Michele all'Adige Fondazione Edmund Mach; Italia
Fil: Khomenko, Iuliia. Instituto Agrario San Michele all'Adige Fondazione Edmund Mach; Italia
Fil: Pedrotti, Michele. Instituto Agrario San Michele all'Adige Fondazione Edmund Mach; Italia
Fil: Romano, Andrea. Instituto Agrario San Michele all'Adige Fondazione Edmund Mach; Italia
Fil: Biasioli, Franco. Instituto Agrario San Michele all'Adige Fondazione Edmund Mach; Italia
Materia
TabPFN
PTR-TOF-MS
VOLATILE ORGANC COMPOUNDS
Nivel de accesibilidad
acceso abierto
Condiciones de uso
https://creativecommons.org/licenses/by-nc-nd/2.5/ar/
Repositorio
CONICET Digital (CONICET)
Institución
Consejo Nacional de Investigaciones Científicas y Técnicas
OAI Identificador
oai:ri.conicet.gov.ar:11336/289545
Acceder
id CONICETDig_17be4414397e08c27cd294b3ce59bb59
oai_identifier_str oai:ri.conicet.gov.ar:11336/289545
network_acronym_str CONICETDig
repository_id_str 3498
network_name_str CONICET Digital (CONICET)
spelling On the use of TabPFN on mass spectrometry analysis of volatile organic compoundsGranitto, Pablo MiguelBetta, EmanuelaKhomenko, IuliiaPedrotti, MicheleRomano, AndreaBiasioli, FrancoTabPFNPTR-TOF-MSVOLATILE ORGANC COMPOUNDShttps://purl.org/becyt/ford/1.2https://purl.org/becyt/ford/1Volatile organic compounds (VOCs) are key markers in applications ranging from food quality assessment to medical diagnostics that can be profiled, for example, by gas chromatography–mass spectrometry (GC-MS) or by direct injection mass spectrometry (e.g. proton transfer reaction mass spectrometry). The common practice in both cases is to construct a tabular dataset from the raw measurements by performing peak extraction across samples and use statistical or machine learning methods to analyze it. However, modeling VOC profiles is particularly challenging due to high dimensionality, noise, and small sample sizes. In this study, we evaluate the Tabular Prior-data Fitted Network (TabPFN), a foundation model recently introduced for tabular data, across diverse VOC datasets. Without requiring task-specific training, TabPFN achieves state-of-the-art performance in both classification and regression tasks, outperforming classical machine learning methods for most datasets. We further explore new strategies to enhance TabPFN’s performance, including ensembling and fine-tuning, finding that a plain ensemble seems to be the best option in this setting. Our results demonstrate that TabPFN is a highly effective modeling tool for VOC profiles obtained with different analytical approaches. It offers robust predictions even in the data-scarce, high-variability scenarios typical of real-world workflows.Fil: Granitto, Pablo Miguel. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Conicet - Rosario. Centro Internacional Franco Argentino de Ciencias de la Información y de Sistemas. Universidad Nacional de Rosario. Centro Internacional Franco Argentino de Ciencias de la Información y de Sistemas; ArgentinaFil: Betta, Emanuela. Instituto Agrario San Michele all'Adige Fondazione Edmund Mach; ItaliaFil: Khomenko, Iuliia. Instituto Agrario San Michele all'Adige Fondazione Edmund Mach; ItaliaFil: Pedrotti, Michele. Instituto Agrario San Michele all'Adige Fondazione Edmund Mach; ItaliaFil: Romano, Andrea. Instituto Agrario San Michele all'Adige Fondazione Edmund Mach; ItaliaFil: Biasioli, Franco. Instituto Agrario San Michele all'Adige Fondazione Edmund Mach; ItaliaNature2025-12info:eu-repo/semantics/articleinfo:eu-repo/semantics/publishedVersionhttp://purl.org/coar/resource_type/c_6501info:ar-repo/semantics/articuloapplication/pdfapplication/pdfapplication/pdfhttp://hdl.handle.net/11336/289545Granitto, Pablo Miguel; Betta, Emanuela; Khomenko, Iuliia; Pedrotti, Michele; Romano, Andrea; et al.; On the use of TabPFN on mass spectrometry analysis of volatile organic compounds; Nature; Scientific Reports; 16; 1; 12-2025; 1-112045-2322CONICET DigitalCONICETenginfo:eu-repo/semantics/altIdentifier/url/https://www.nature.com/articles/s41598-025-29128-6info:eu-repo/semantics/altIdentifier/doi/10.1038/s41598-025-29128-6info:eu-repo/semantics/openAccesshttps://creativecommons.org/licenses/by-nc-nd/2.5/ar/reponame:CONICET Digital (CONICET)instname:Consejo Nacional de Investigaciones Científicas y Técnicas2026-08-25T14:53:49Zoai:ri.conicet.gov.ar:11336/289545instacron:CONICETInstitucionalhttp://ri.conicet.gov.ar/Organismo científico-tecnológicoNo correspondehttp://ri.conicet.gov.ar/oai/requestdasensio@conicet.gov.ar; lcarlino@conicet.gov.arArgentinaNo correspondeNo correspondeNo correspondeopendoar:34982026-08-25 14:53:49.697CONICET Digital (CONICET) - Consejo Nacional de Investigaciones Científicas y Técnicasfalse
dc.title.none.fl_str_mv On the use of TabPFN on mass spectrometry analysis of volatile organic compounds
title On the use of TabPFN on mass spectrometry analysis of volatile organic compounds
spellingShingle On the use of TabPFN on mass spectrometry analysis of volatile organic compounds
Granitto, Pablo Miguel
TabPFN
PTR-TOF-MS
VOLATILE ORGANC COMPOUNDS
title_short On the use of TabPFN on mass spectrometry analysis of volatile organic compounds
title_full On the use of TabPFN on mass spectrometry analysis of volatile organic compounds
title_fullStr On the use of TabPFN on mass spectrometry analysis of volatile organic compounds
title_full_unstemmed On the use of TabPFN on mass spectrometry analysis of volatile organic compounds
title_sort On the use of TabPFN on mass spectrometry analysis of volatile organic compounds
dc.creator.none.fl_str_mv Granitto, Pablo Miguel
Betta, Emanuela
Khomenko, Iuliia
Pedrotti, Michele
Romano, Andrea
Biasioli, Franco
author Granitto, Pablo Miguel
author_facet Granitto, Pablo Miguel
Betta, Emanuela
Khomenko, Iuliia
Pedrotti, Michele
Romano, Andrea
Biasioli, Franco
author_role author
author2 Betta, Emanuela
Khomenko, Iuliia
Pedrotti, Michele
Romano, Andrea
Biasioli, Franco
author2_role author
author
author
author
author
dc.subject.none.fl_str_mv TabPFN
PTR-TOF-MS
VOLATILE ORGANC COMPOUNDS
topic TabPFN
PTR-TOF-MS
VOLATILE ORGANC COMPOUNDS
purl_subject.fl_str_mv https://purl.org/becyt/ford/1.2
https://purl.org/becyt/ford/1
dc.description.none.fl_txt_mv Volatile organic compounds (VOCs) are key markers in applications ranging from food quality assessment to medical diagnostics that can be profiled, for example, by gas chromatography–mass spectrometry (GC-MS) or by direct injection mass spectrometry (e.g. proton transfer reaction mass spectrometry). The common practice in both cases is to construct a tabular dataset from the raw measurements by performing peak extraction across samples and use statistical or machine learning methods to analyze it. However, modeling VOC profiles is particularly challenging due to high dimensionality, noise, and small sample sizes. In this study, we evaluate the Tabular Prior-data Fitted Network (TabPFN), a foundation model recently introduced for tabular data, across diverse VOC datasets. Without requiring task-specific training, TabPFN achieves state-of-the-art performance in both classification and regression tasks, outperforming classical machine learning methods for most datasets. We further explore new strategies to enhance TabPFN’s performance, including ensembling and fine-tuning, finding that a plain ensemble seems to be the best option in this setting. Our results demonstrate that TabPFN is a highly effective modeling tool for VOC profiles obtained with different analytical approaches. It offers robust predictions even in the data-scarce, high-variability scenarios typical of real-world workflows.
Fil: Granitto, Pablo Miguel. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Conicet - Rosario. Centro Internacional Franco Argentino de Ciencias de la Información y de Sistemas. Universidad Nacional de Rosario. Centro Internacional Franco Argentino de Ciencias de la Información y de Sistemas; Argentina
Fil: Betta, Emanuela. Instituto Agrario San Michele all'Adige Fondazione Edmund Mach; Italia
Fil: Khomenko, Iuliia. Instituto Agrario San Michele all'Adige Fondazione Edmund Mach; Italia
Fil: Pedrotti, Michele. Instituto Agrario San Michele all'Adige Fondazione Edmund Mach; Italia
Fil: Romano, Andrea. Instituto Agrario San Michele all'Adige Fondazione Edmund Mach; Italia
Fil: Biasioli, Franco. Instituto Agrario San Michele all'Adige Fondazione Edmund Mach; Italia
description Volatile organic compounds (VOCs) are key markers in applications ranging from food quality assessment to medical diagnostics that can be profiled, for example, by gas chromatography–mass spectrometry (GC-MS) or by direct injection mass spectrometry (e.g. proton transfer reaction mass spectrometry). The common practice in both cases is to construct a tabular dataset from the raw measurements by performing peak extraction across samples and use statistical or machine learning methods to analyze it. However, modeling VOC profiles is particularly challenging due to high dimensionality, noise, and small sample sizes. In this study, we evaluate the Tabular Prior-data Fitted Network (TabPFN), a foundation model recently introduced for tabular data, across diverse VOC datasets. Without requiring task-specific training, TabPFN achieves state-of-the-art performance in both classification and regression tasks, outperforming classical machine learning methods for most datasets. We further explore new strategies to enhance TabPFN’s performance, including ensembling and fine-tuning, finding that a plain ensemble seems to be the best option in this setting. Our results demonstrate that TabPFN is a highly effective modeling tool for VOC profiles obtained with different analytical approaches. It offers robust predictions even in the data-scarce, high-variability scenarios typical of real-world workflows.
publishDate 2025
dc.date.none.fl_str_mv 2025-12
dc.type.none.fl_str_mv info:eu-repo/semantics/article
info:eu-repo/semantics/publishedVersion
http://purl.org/coar/resource_type/c_6501
info:ar-repo/semantics/articulo
format article
status_str publishedVersion
dc.identifier.none.fl_str_mv http://hdl.handle.net/11336/289545
Granitto, Pablo Miguel; Betta, Emanuela; Khomenko, Iuliia; Pedrotti, Michele; Romano, Andrea; et al.; On the use of TabPFN on mass spectrometry analysis of volatile organic compounds; Nature; Scientific Reports; 16; 1; 12-2025; 1-11
2045-2322
CONICET Digital
CONICET
url http://hdl.handle.net/11336/289545
identifier_str_mv Granitto, Pablo Miguel; Betta, Emanuela; Khomenko, Iuliia; Pedrotti, Michele; Romano, Andrea; et al.; On the use of TabPFN on mass spectrometry analysis of volatile organic compounds; Nature; Scientific Reports; 16; 1; 12-2025; 1-11
2045-2322
CONICET Digital
CONICET
dc.language.none.fl_str_mv eng
language eng
dc.relation.none.fl_str_mv info:eu-repo/semantics/altIdentifier/url/https://www.nature.com/articles/s41598-025-29128-6
info:eu-repo/semantics/altIdentifier/doi/10.1038/s41598-025-29128-6
dc.rights.none.fl_str_mv info:eu-repo/semantics/openAccess
https://creativecommons.org/licenses/by-nc-nd/2.5/ar/
eu_rights_str_mv openAccess
rights_invalid_str_mv https://creativecommons.org/licenses/by-nc-nd/2.5/ar/
dc.format.none.fl_str_mv application/pdf
application/pdf
application/pdf
dc.publisher.none.fl_str_mv Nature
publisher.none.fl_str_mv Nature
dc.source.none.fl_str_mv reponame:CONICET Digital (CONICET)
instname:Consejo Nacional de Investigaciones Científicas y Técnicas
reponame_str CONICET Digital (CONICET)
collection CONICET Digital (CONICET)
instname_str Consejo Nacional de Investigaciones Científicas y Técnicas
repository.name.fl_str_mv CONICET Digital (CONICET) - Consejo Nacional de Investigaciones Científicas y Técnicas
repository.mail.fl_str_mv dasensio@conicet.gov.ar; lcarlino@conicet.gov.ar
_version_ 1874774846287642624
score 13.265058

Publicaciones similares