Data-handling strategies for metabonomic studies: example of the UHPLC-ESI/ToF urinary signature of tetrahydrocannabinol in humans

Abstract : Metabonomics has become a very valuable tool and many research fields rely on results coming out from this combination of analytical techniques, chemometric strategies, and biological interpretation. Moreover, the matrices are more and more complex and the implications of the results are often of major importance. In this context, the need for pertinent validation strategies comes naturally. The choice of the appropriate chemometric method remains nevertheless a difficult task due to particularities such as: the number of measured variables, the complexity of the matrix and the purposes of the study. Consequently, this paper presents a detailed metabonomic study on human urine with a special emphasis on the importance of assessing the data's quality. It also describes, step by step, the statistical tools currently used and offers a critical view on some of their limits. In this work, 29 urine samples among which 15 samples obtained from tetrahydrocannabinol (delta-9-tetrahydrocannabinol)-consuming athletes, 5 samples provided by volunteers, and 9 samples obtained from athletes were submitted to untargeted analysis by means of ultra high-pressure liquid chromatography-electrospray ionization-time-of-flight mass spectrometry. Next, the quality of the obtained data was assessed and the results were compared to those found in databases. Then, unsupervised (principal component analysis (PCA)) and supervised (ANOVA/PCA, partial least-square-discriminant analysis (PLS-DA), orthogonal PLS-DA) univariate and multivariate statistical methods were applied.
Type de document :
Article dans une revue
Analytical and Bioanalytical Chemistry, Springer Verlag, 2014, 406 (4), pp.1209-1219. 〈10.1007/s00216-013-7199-0〉
Liste complète des métadonnées

https://hal.archives-ouvertes.fr/hal-00850196
Contributeur : Agnès Bussy <>
Soumis le : lundi 5 août 2013 - 16:10:02
Dernière modification le : jeudi 8 février 2018 - 11:10:19

Identifiants

Collections

Citation

Agnèta Kiss, Claire Bordes, Corinne Buisson, Françoise Lasne, Pierre Lanteri, et al.. Data-handling strategies for metabonomic studies: example of the UHPLC-ESI/ToF urinary signature of tetrahydrocannabinol in humans. Analytical and Bioanalytical Chemistry, Springer Verlag, 2014, 406 (4), pp.1209-1219. 〈10.1007/s00216-013-7199-0〉. 〈hal-00850196〉

Partager

Métriques

Consultations de la notice

149