incubatoR

Description:

This repository is containing the code used in "Improving the screening analysis of pesticide metabolites in human biomonitoring by combining high-throughput in-vitro incubation and automated LC-HRMS data processing" for reproducibility.

Main task is to identify metabolic transformation products in a LC-HRMS2 dataset of in-parallel incubated xenobiotic compounds by applying statistical tools as well as mass defect filtering and extraction of the respective mass spectra for spectal library generation.

An example dataset has been deposited to the EMBL-EBI MetaboLights database (DOI: 10.1093/nar/gkz1019, PMID:31691833) with the identifier MTBLS2402 and is accessible directly at https://www.ebi.ac.uk/metabolights/MTBLS2402.

The code was modified to run on a CentOS7 high performance computation (HPC) cluster based on the R language (v 3.6.1) and wrapped in bash shell commands for parallel processing and job submission using the unix shell job sheduler command qsub.

Sample nomenclature:

The sample set should contain measurements of incubated replicates of each compound, a reference standard solution, negative controls and injection/ sample peparation blanks.

The incubated sample names contain a string identifing the different xenobiotic compounds used for incubation (e.g. Terbuthylazine) followed by a _R and a numeric value idenfing the replicate (e.g. XX_Terbuthylazine_R1.mzML).

The reference standard solution names contain the name of the xenobiotic compound followed by a sting identifing to be a reference standard solution (e.g. _clean XX_Terbuthylazine_clean.mzML).

The negative controls contain a string intentifing to be a negative control (e.g. XX_NC1.mzML).

Sample extraction blanks and Injection blanks should also be identified by a unique string (e.g. XX_Blank1.mzML or XX_B.mzML).

Different ionization modes are stored in seperate folders named Pos and Neg.

Order of calculation steps:

Follwing calculation steps are provided:

1. Peaklist generation (XCMS [1] and CAMERA [2]) by Rscripts/xcms.R and Rscripts/camera.R (bash/jobsubmit_1xcms.sh and bash/jobsubmit_1camera.sh for parallel job submission).

   XCMS:

   INPUT: settings_xcms.yaml, class.csv, globalvar.sh

   OUTPUT: xcms.rds

   CAMERA:

   INPUT: settings_camera.yaml, xcms.rds, globalvar.sh

   OUTPUT: camera.rds, metadata.tsv, peaklist.tsv

2. Calculation of the statistical comparisson by Rscripts/statistics.R (bash/jobsubmit_2statistics.sh for parallel job submission), including the package Rvolcano [3] in case of the application of robust stastistics.

   INPUT: class.csv, metadata.tsv, peaklist.tsv, settings_statistic.yaml, globalvar.sh

   OUTPUT: compound/Result_compound.rds

3. Filtering of non-metabolic features by several cut-off values and plotting for manual evaluation by Rscripts/metabolites.R (bash/jobsubmit_3metabolites.sh for parallel job submission).

   INPUT: class.csv, Result_compound.rds, rt_compound.txt, parent_compound.txt, target_compound.txt, settings_metabolites.yaml, globalvar.sh

   OUTPUT: compound/diffplot.png, compound/volcano.png, compound/MDF.png, compound/Feature.png, compound/Metabolites.txt

4. EIC extraction of the suspected metabolite featuresRscripts/eic.R (bash/jobsubmit_4eic.sh for parallel job submission), based on MSnBase [4].

   INPUT: class.csv, compound/Metabolites.txt, globalvar.sh

   OUTPUT: compound/EIC_metabolite

5. MSMS extraction an spectral purity evaluation (implementation of MSpurity [5]) by Rscripts/ddextract.R (bash/jobsubmit_5ddextract.sh for parallel job submission).

   INPUT: class.csv, compound/Metabolites.txt, parameter_msms.sh, globalvar.sh

   OUTPUT: compound/MSMS/*

6. Molecular formula annotation - implementation of GenForm [6] (bash/jobsubmit_6genform.sh for parallel job submission).

   INPUT: class.csv, compound/MSMS/*, compound/MSMS/*/MS1.txt, FF_compound.txt, parameter_genform.sh, globalvar.sh

   OUTPUT: compound/MSMS/*/*.out compound/MSMS/*/Clean_*.txt

Description of input files:

globalvar.sh: environmental variables used for running the scripts

class.csv: class file used in xcms

settings_xcms.yaml: xcms settings

settings_camera.yaml: camera settings

settings_statistics.yaml: parameters applied in the calculation of the statistics

settings_metabolites.yaml: parameters applied in the metabolite filtering

parameter_genform.sh: parameters applied in the application of GenForm

parent_compound.txt: m/z value of the predicted ion for the parent compound

target_compound.txt: list of m/z values of already known metabolites

RT_compound.txt: Retention time of the parent compound

FF_compound.txt: Fuzzy formula applied in GenForm molecular formula annotation

Description of output files:

xcms.rds: RData file containing the xcmsSet

camera.rds: RData file containing the ouput of CAMERA

metadata.tsv: tab seperated metadata of CAMERA

peaklist.tsv:tab seperated peaklist

compound/Result_compound.rds: RData file containing the calculated statistics

compound/diffplot.png: Visualization of the fold change cut-off, representing the log2 of the mean intensity of the incubated group over the mean of the other samples

compound/volcano.png: Visualization of the p-value over the log2 fold change

compound/MDF.png: Visualization of the mass defect shift (compared to the parent compound) and m/z value

compound/Feature.png: Visualization the m/z value over retention time of all remaining features

compound/Metabolite.txt: Text file containing the remaining feature peakids (mass@rt)

compound/EIC: Folder containing the EIC of all features - color coded by group

compound/MSMS/*: Folder containing the extracted MSMS spectra as .txt for all applied features

compound/MSMS/*/*.out : Genform output for all applied MSMS spectra

compound/MSMS/*/Clean_*.txt: Spectra only containing the fragments that can be explained by the given formula

References:

[1] Smith, C. A.; Want, E. J.; O’Maille, G.; Abagyan, R.; Siuzdak, G. XCMS: Processing Mass Spectrometry Data for Metabolite Profiling Using Nonlinear Peak Alignment, Matching, and Identification. Anal. Chem. 2006, 78 (3), 779–787. https://doi.org/10.1021/ac051437y.

[2] Hochreiter, S. Bioinformatics Research and Development: First International Conference, BIRD 2007, Berlin, Germany, March 12-14, 2007, Proceedings; Springer Science & Business Media, 2007.

[3] Kumar, N.; Hoque, Md. A.; Sugimoto, M. Robust Volcano Plot: Identification of Differential Metabolites in the Presence of Outliers. BMC Bioinformatics 2018, 19. https://doi.org/10.1186/s12859-018-2117-2.

[4] Gatto, L.; Lilley, K. S. MSnbase-an R/Bioconductor Package for Isobaric Tagged Mass Spectrometry Data Visualization, Processing and Quantitation. Bioinformatics 2012, 28 (2), 288–289. https://doi.org/10.1093/bioinformatics/btr645.

[5] Lawson, T. N.; Weber, R. J. M.; Jones, M. R.; Chetwynd, A. J.; Rodrı́guez-Blanco, G.; Di Guida, R.; Viant, M. R.; Dunn, W. B. MsPurity: Automated Evaluation of Precursor Ion Purity for Mass Spectrometry-Based Fragmentation in Metabolomics. Anal. Chem. 2017, 89 (4), 2432–2439. https://doi.org/10.1021/acs.analchem.6b04358.

[6] Meringer, M.; Reinker, S.; Zhang, J.; Muller, A. MS/MS Data Improves Automated Determination of Molecular Formulas by Mass Spectrometry. MATCH Communications in Mathematical and in Computer Chemistry 2011, 65, 259–290.