Untargeted time-pattern analysis of LC-HRMS data to detect spills and compounds with high fluctuation in influent wastewater

Peak prioritization plays a key role in non-target analysis of complex samples in order to focus the elucidation efforts on potentially relevant substances. The present work shows the development of a computational workflow capable of detecting compounds that exhibit large variation in intensity ove...

Descripción completa

Detalles Bibliográficos
Autores: Alygizakis, Nikiforos A., Gago-Ferrero, Pablo, Hollender, Juliane, Thomaidis, Nikolaos S.
Tipo de recurso: artículo
Estado:Versión aceptada para publicación
Fecha de publicación:2019
País:España
Institución:Consejo Superior de Investigaciones Científicas (CSIC)
Repositorio:DIGITAL.CSIC. Repositorio Institucional del CSIC
OAI Identifier:oai:digital.csic.es:10261/345524
Acceso en línea:http://hdl.handle.net/10261/345524
https://api.elsevier.com/content/abstract/scopus_id/85052503605
Access Level:acceso abierto
Palabra clave:Wastewater
Multivariate Empirical Bayes Approach
Non-target screening
Peak prioritization workflow
Trend analysis
http://metadata.un.org/sdg/6
http://metadata.un.org/sdg/3
Ensure healthy lives and promote well-being for all at all ages
Ensure availability and sustainable management of water and sanitation for all
Descripción
Sumario:Peak prioritization plays a key role in non-target analysis of complex samples in order to focus the elucidation efforts on potentially relevant substances. The present work shows the development of a computational workflow capable of detecting compounds that exhibit large variation in intensity over time. The developed approach is based on three open-source R packages (xcms, CAMERA and TIMECOURSE) and includes the use of the statistical test Multivariate Empirical Bayes Approach to rank the compounds based on the Hotelling T2 coefficient, which is an indicator of large concentration variations of unknown components. The approach was applied to replicate series of 24 h composite flow-proportional influent wastewater samples collected during 8 consecutive days. 60 events involving unknown substances with high fluctuation over time were successfully prioritized. 14 of those compounds were tentatively identified using HRMS/MS libraries, chemical databases, in-silico fragmentation tools, and retention time prediction models. Four compounds were confirmed with standards from which two never reported before in wastewater.