Powerful multiple testing procedure with application in metabolomics
Rémi Servien

In metabolomics, each complex mixture of interest gives a spectrum that could be viewed as a functional data. This spectrum is the weighted combination of all the spectra of the metabolites that composed the complex mixture. One of the scientific question is then to derive the composition of the complex mixture from its spectrum and a database with all the spectra of the metabolites. To achieve this goal, we develop a new powerful multiple testing procedure, based on the maximum likelihood estimator, that control the Family Wise Error Rate (FWER). Numerical and real data experiments highlight the performance of our approach compared to the state-of-the-art procedures.
P. Tardivel, R. Servien and D. Concordet. A powerful multiple testing procedure in linear Gaussian model. Preprint. https://hal.archives-ouvertes.fr/hal-01322077
P. Tardivel, C. Canlet, G. Lefort, M. Tremblay-Franco, L. Debrauwer, D. Concordet and R. Servien. ASICS : an automatic method for identi\u000Ccation and quanti\u000Ccation of metabolites in complex 1D 1H NMR spectra, Metabolomics (2017), 13(10), 109, https://hal.archives-ouvertes.fr/hal-01535613."10:30-11:00 - R.Servien