A machine learning approach for handling big data produced by high resolution mass spectrometry after data independent acquisition of small molecules - Proof of concept study using an artificial neural network for sample classification.
Big Data
Chromatography, Liquid
Cocaine
/ blood
Humans
Machine Learning
Mass Spectrometry
/ statistics & numerical data
Neural Networks, Computer
Proof of Concept Study
Reproducibility of Results
Sensitivity and Specificity
Small Molecule Libraries
Substance Abuse Detection
/ methods
Zolpidem
/ blood
SWATH
data independent acquisition
high resolution mass spectrometry
machine learning
small molecules
Journal
Drug testing and analysis
ISSN: 1942-7611
Titre abrégé: Drug Test Anal
Pays: England
ID NLM: 101483449
Informations de publication
Date de publication:
Jun 2020
Jun 2020
Historique:
received:
08
07
2019
revised:
17
01
2020
accepted:
28
01
2020
pubmed:
31
1
2020
medline:
1
4
2021
entrez:
31
1
2020
Statut:
ppublish
Résumé
Liquid chromatography coupled to high-resolution mass spectrometry (HRMS) enables data independent acquisition (DIA) and untargeted screening. However, to avoid the handling of the resulting large dataset, most laboratories in that field still use targeted screening methods, which offer good sensitivity and specificity but are limited to known compounds. The promising field of machine learning offers new possibilities such as artificial neural networks that can be trained to classify large amounts of data. In this proof of concept study, we exemplify such a machine learning approach for raw HRMS-DIA data files. We evaluated a machine learning model using training, validation, and test sets of solvent and whole blood samples containing drugs (of abuse) common in forensic toxicology. For that purpose, different platforms were used. With a feedforward neural network model architecture, a category prediction (blank sample vs. drug containing sample) was aimed for. With the applied machine learning approaches, the sensitivity and specificity, of the validation and test set, for the prediction of sample classes were in a suitable range for an actual use in a (routine) laboratory (e.g. workplace drug testing). In conclusion, this proof of concept study clearly demonstrated the huge potential of machine learning in the analysis of HRMS-DIA data.
Substances chimiques
Small Molecule Libraries
0
Zolpidem
7K383OQI23
Cocaine
I5Y540LHVR
Types de publication
Journal Article
Langues
eng
Sous-ensembles de citation
IM
Pagination
836-845Informations de copyright
© 2020 John Wiley & Sons, Ltd.
Références
Marquet P. LC-MS vs. GC-MS, Online extraction systems, advantages of technology for drug screening assays. In: Langman LJ, Snozek CLH, eds. LC-MS in drug analysis: Methods and protocols. Totowa, NJ: Humana Press; 2012:15-27.
Wu AH, Gerona R, Armenian P, French D, Petrie M, Lynch KL. Role of liquid chromatography-high-resolution mass spectrometry (LC-HR/MS) in clinical toxicology. Clin Toxicol. 2012;50(8):733-742.
Gómez MJ, Gómez-Ramos MM, Malato O, Mezcua M, Férnandez-Alba AR. Rapid automated screening, identification and quantification of organic micro-contaminants and their main transformation products in wastewater and river waters using liquid chromatography-quadrupole-time-of-flight mass spectrometry with an accurate-mass database. J Chromatogr A. 2010;1217(45):7038-7054.
Mollerup CB, Dalsgaard PW, Mardal M, Linnet K. Targeted and non-targeted drug screening in whole blood by UHPLC-TOF-MS with data-independent acquisition. Drug Test Anal. 2017;9(7):1052-1061.
Maurer HH. What is the future of (ultra) high performance liquid chromatography coupled to low and high resolution mass spectrometry for toxicological drug screening? J Chromatogr A. 2013;1292:19-24.
Arnhard K, Gottschall A, Pitterl F, Oberacher H. Applying ‘sequential windowed acquisition of all theoretical fragment ion mass spectra' (SWATH) for systematic toxicological analysis with liquid chromatography-high-resolution tandem mass spectrometry. Anal Bioanal Chem. 2015;407(2):405-414.
Roemmelt AT, Steuer AE, Kraemer T. Liquid chromatography, in combination with a quadrupole time-of-flight instrument, with sequential window acquisition of all theoretical fragment-ion spectra acquisition: validated quantification of 39 antidepressants in whole blood as part of a simultaneous screening and quantification procedure. Anal Chem. 2015;87(18):9294-9301.
Heywood D. SONAR - delivering MS/MS data from a DIA experiment. Waters White Paper 2017.
Sundström M, Pelander A, Ojanperä I. Comparison of post-targeted and pre-targeted urine drug screening by UHPLC-HR-QTOFMS. J Anal Toxicol. 2017;41(7):623-630.
Pasin D, Bidny S, Fu S. Analysis of new designer drugs in post-mortem blood using high-resolution mass spectrometry. J Anal Toxicol. 2015;39(3):163-171.
McAlister G, Ntai I, Kiyonami R, et al. New method filters for improved MSn acquisition for small molecule and proteomics workflows. San Jose, CA: Poster from Thermo Fisher Scientific; 2018.
Boxler MI, Streun GL, Liechti ME, Schmid Y, Kraemer T, Steuer AE. Human metabolome changes after a single dose of 3,4-methylenedioxymethamphetamine (MDMA) with special focus on steroid metabolism and inflammation processes. J Proteome Res. 2018;17(8):2900-2907.
Elmiger MP, Poetzsch M, Steuer AE, Kraemer T. Parameter optimization for feature and hit generation in a general unknown screening method - proof of concept study using a Design of Experiment Approach for a high resolution mass spectrometry procedure after data independent acquisition. Anal Chem. 2018;90(5):3531-3536.
Bengio Y, Courville A, Vincent P. Representation learning: a review and new perspectives. IEEE Trans Pattern Anal Mach Intell. 2013;35(8):1798-1828.
Zhou D, Miao L, He Y. Position-aware deep multi-task learning for drug-drug interaction extraction. Artif Intell Med. 2018;87:1-8.
Lee J-G, Jun S, Cho Y-W, et al. Deep learning in medical imaging: general overview. Korean J Radiol. 2017;18(4):570-584.
LeCun Y, Bengio Y, Hinton G. Deep learning. Nature. 2015;521:436.
Choy G, Khalilzadeh O, Michalski M, et al. Current applications and future impact of machine learning in radiology. Radiology. 2018;288(2):318-328.
Brouard C, Rousu J, Szedmak S, Bach E, Böcker S. Liquid-chromatography retention order prediction for metabolite identification. Bioinformatics. 2018;34(17):i875-i883.
Dührkop K, Shen H, Meusel M, Rousu J, Böcker S. Searching molecular structure databases with tandem mass spectra using CSI:FingerID. Proc Natl Acad Sci. 2015;112(41):12580-12585.
van der Hooft JJJ, Wandy J, Barrett MP, Burgess KEV, Rogers S. Topic modeling for untargeted substructure exploration in metabolomics. Proc Natl Acad Sci. 2016;113(48):13738-13743.
Berthold MR, Cebron N, Dill F, et al. KNIME: The Konstanz Information Miner. 2008; Berlin, Heidelberg: Springer Berlin Heidelberg.
François Chollet. Keras. GitHub, GitHub repository; 2015.
Elmiger MP, Poetzsch M, Steuer AE, Kraemer T. Assessment of simpler calibration models in the development and validation of a fast postmortem multi-analyte LC-QTOF quantitation method in whole blood with simultaneous screening capabilities using SWATH acquisition. Anal Bioanal Chem. 2017;409(27):6495-6508.
Team RDC. R: A Language and Environment for Statistical Computing. 2018; https://www.R-project.org
Team R. RStudio: Integrated Development for R. RStudio, Inc., Boston, MA. 2016; http://www.rstudio.com/
Nawi NM, Atomi WH, Rehman MZ. The effect of data pre-processing on optimized training of artificial neural networks. Procedia Technology. 2013;11(Complete):32-39.
Kumar P. (2009). Use of Fuzzy Set and Neural Network to Extract Fingerprint Minutiae Points and Location (master's thesis). Retrieved from https://www.semanticscholar.org
Skymind. A Beginner's Guide to LSTMs and Recurrent Neural Networks. https://skymind.ai/wiki/lstm. Accessed 15th Dec 2019.
Galetzka M. (2014). Intelligent Predictions: an empirical study of the Cortical Learning Algorithm (master's thesis). Retrieved from https://www.semanticscholar.org