Pré-Publication, Document De Travail Année : 2021

Algorithmic Learning for Auto-deconvolution of GC-MS Data to Enable Molecular Networking within GNPS

Alexander Aksenov (1, 2) , Ivan Laponogov (3) , Zheng Zhang (1) , Sophie Lf Doran (3) , Ilaria Belluomo (3) , Dennis Veselkov , Wout Bittremieux (1, 2, 4) , Louis Felix Nothias (1, 2) , Mélissa Nothias-Esposito (1, 2) , Katherine Maloney (1) , Biswapriya Misra (5) , Alexey Melnik (1) , Kenneth Jones (1) , Kathleen Dorrestein (1, 2) , Morgan Panitchpakdi (1) , Madeleine Ernst (1, 6) , Justin J.J. van Der Hooft (1, 7) , Mabel Gonzalez (8) , Chiara Carazzone (8) , Adolfo Amézquita (8) , Chris Callewaert (9, 10) , James Morton (10) , Robert Quinn (11) , Amina Bouslimani (1, 2) , Andrea Albarracín Orio (12) , Daniel Petras (1, 2) , Andrea Smania (13) , Sneha Couvillion (14) , Meagan Burnet (14) , Carrie Nicora (14) , Erika Zink (14) , Thomas Metz (14) , Viatcheslav Artaev (15) , Elizabeth Humston-Fulmer (15) , Rachel Gregor (16) , Michael Meijler (16) , Itzhak Mizrahi (16) , Stav Eyal (16) , Brooke Anderson (17) , Rachel Dutton (17) , Raphaël Lugan (18) , Pauline Le Boulch (18) , Yann Guitton (19) , Stéphanie Prévost (19) , Audrey Poirier (19) , Gaud Dervilly (19) , Bruno Le Bizec (19) , Aaron Fait (16) , Noga Sikron Persi (16) , Chao Song (16) , Kelem Gashu (16) , Roxana Coras (4) , Monica Guma (4) , Julia Manasson (20) , Jose Scher (20) , Dinesh Barupal (21) , Saleh Alseekh (22, 23) , Alisdair Fernie (22, 23) , Reza Mirnezami (24) , Vasilis Vasiliou (25) , Robin Schmid (26) , Roman Borisov (27) , Larisa Kulikova (28) , Rob Knight (4) , Mingxun Wang (1, 2) , George Hanna (3) , Pieter Dorrestein (1, 2, 4) , Kirill Veselkov (3)
1 Skaggs School of Pharmacy and Pharmaceutical Sciences [San Diego]
2 Collaborative Mass Spectrometry Innovation Center
3 Imperial College London
4 UC San Diego - University of California [San Diego]
5 Wake Forest School of Medicine [Winston-Salem]
6 Statens Serum Institut [Copenhagen]
7 WUR - Wageningen University and Research [Wageningen]
8 UNIANDES - Universidad de los Andes [Bogota]
9 CMET - Center for Microbial Ecology and Technology
10 UC - University of California
11 Michigan State University [East Lansing]
12 UCC - Universidad Católica de Córdoba = Catholic University of Córdoba
13 Universidad Nacional de Córdoba [Argentina]
14 PNNL - Pacific Northwest National Laboratory
15 LECO Corporation
16 BGU - Ben-Gurion University of the Negev
17 USD - University of San Diego
18 UMR QualiSud - Démarche intégrée pour l'obtention d'aliments de qualité
19 LABERCA - Laboratoire d'étude des Résidus et Contaminants dans les Aliments
20 NYU - New York University School of Medicine
21 West Coast Metabolomics Center
22 MPI-MP - Max Planck Institute of Molecular Plant Physiology
23 CPSBB - Center of Plant Systems Biology and Biotechnology [Plovdiv]
24 NHS Foundation Trust [London]
25 YSPH - Yale School of Public Health
26 Institute of Inorganic and Analytical chemistry [University of Freiburg] | Institut für Anorganische und Analytische Chemie [Albert-Ludwigs-Universität Freiburg]
27 TIPS - A.V. Topchiev Institute of Petrochemical Synthesis
28 RUDN - Université russe de l'amitié des peuples = People's Friendship University of Russia = Rossijskij universitet družby narodov [Moscou]
Dennis Veselkov
  • Fonction : Auteur

Résumé

Gas chromatography-mass spectrometry (GC-MS) represents an analytical technique with significant practical societal impact. Spectral deconvolution is an essential step for interpreting GC-MS data. No public GC-MS repositories that also enable repository-scale analysis exist, in part because deconvolution requires significant user input. We therefore engineered a scalable machine learning workflow for the Global Natural Product Social Molecular Networking (GNPS) analysis platform to enable the mass spectrometry community to store, process, share, annotate, compare, and perform molecular networking of GC-MS data. The workflow performs auto-deconvolution of compound fragmentation patterns via unsupervised non-negative matrix factorization, using a Fast Fourier Transform-based strategy to overcome scalability limitations. We introduce a “balance score” that quantifies the reproducibility of fragmentation patterns across all samples. We demonstrate the utility of the platform with breathomics analysis applied to the early detection of oesophago-gastric cancer, and by creating the first molecular spatial map of the human volatilome.

Dates et versions

hal-03173091 , version 1 (18-03-2021)

Identifiants

Citer

Alexander Aksenov, Ivan Laponogov, Zheng Zhang, Sophie Lf Doran, Ilaria Belluomo, et al.. Algorithmic Learning for Auto-deconvolution of GC-MS Data to Enable Molecular Networking within GNPS. 2021. ⟨hal-03173091⟩
210 Consultations
0 Téléchargements

Altmetric

Partager

  • More