Offres & annonces

PhD Computational metabolomics for studying oomycete plant infestation, LC-MS/MS

Main supervisor: Prof. Sebastian Böcker, Bioinformatics, Institute of Mathematics and Informatics, FSU 
Co-supervisor(s): Dr. Ales Svatos, Research Group Mass Spectrometry/Proteomics, MPI-CE 

Project description: LC-MS/MS (Liquid Chromatography – Tandem Mass Spectrometry) is the predominant experimental setup for proteomics experiments, and has also become a prevalent setup in metabolomics. Proteomics has successfully established workflows where identification and even quantification of proteins is mostly based on tandem mass spectrometry (MS/MS) data. In contrast, metabolomics is often carried out at the MS1 level. Even "untargeted" metabolomics experiments often rely on exact monoisotopic masses to identify unknowns; this approach is, by design, highly limited in its power to distinguish compounds.

The Böcker group has recently introduced a method to search MS/MS data of an unknown compound in molecular structure databases such as PubChem, ChemSpider, or ChEBI. The method reaches 150% more correct identification than the second-best methods available for this task (FingerID, CFM-ID), and 250% more correct identifications than MetFrag (currently, the most widely used method). Our method CSI:FingerID will be made available on our website shortly; the corresponding publication is currently in revision with PNAS.

In this project, we want to shift untargeted metabolomics further to the tandem MS side, where fragmentation spectra are the main source of identification, and this information is complemented by MS1 data. Similar approaches have been very successful for the analysis of particular metabolites such as non-ribosomal peptides; see the dereplication networks developed by Pieter Dorrestein's group at UCSD (for example, Watrous et al., PNAS 2012). On the other hand, we want to integrate MS1 data in a more formal way to establish connections between (potentially unknown) molecules: For example, Morreel et al. (Plant Cell, 2014) introduced a promising method to derive hypothetical molecular networks from MS1 and tandem MS data, then confirmed hundreds of novel metabolites by manual and extremely time-intense analysis of tandem MS and MSn data. Similar to the CSI:FingerID search engine, we want to replace this time-intense workflow, at least in part, by automated methods; at the same time, we want to guide experimentalists to the (potentially) most promising compounds and connections, as suggested by the data.

In Ales’ Svatos group at the MPI for Chemical Ecology, a novel method for in situ metabolite profiling with plant-cell resolution is being developed. The system uses a laser ablation electrospray (LAESI) method for metabolite ionization. Formed ions are detected using a hybrid mass spectrometer Synapt (Waters) with ion mobility separation unit. This system will allow us to study fast metabolic responses of plant cells after inoculation with different pathogenic bacteria and fungi. Our method for metabolite identification will be used to follow metabolic changes in the infected cells.

We are searching for a highly motivated PhD student with a babackground in Bioinformatics for this project, experience in chemistry is highly desirable.

1ères JS Numériques du RFMF

Prochain congrès: en Juin 2020, le RFMF organise ses premières JS numériques!

Réseau Francophone de Métabolomique et Fluxomique