Pipeline for the compilations of MassBank records from raw data

Information to be supplied

A meta data spreadsheet containing information for the measured substances

Sheet Compounds

Sheet for the meta data of the sample, compound, spectrum, and authors. The table contains one row for each spectrum.

• File MSP file containing the spectrum; located under path raw data/exported as raw msp in the project folder (mandatory, e.g. Rutin.msp)
• Name Preferred compound name (mandatory, e.g. Rutin)
• Synonyms Common compound name synonyms (optional, e.g. Rutinoside; Eldrin)
• Structure / Unique ID The compound structure - one of the following three options (mandatory, e.g. 5280805)
  • InChI
  • SMILES
  • PubChem CID
• Formula Molecular formula of the compound (optional, e.g. C27H30O16)
• Exact mass The exact mass of the compound (optional, e.g. 610.15339)
• CHROMATOGRAPHY The name of the chromatography instrumentation (mandatory, e.g. Symmetry C18 Column, Waters)
• RT (min) The retention time of the parent ion (mandatory, e.g. 9.5)
• INSTRUMENT The name of the mass spectrometry instrumentation (mandatory, e.g. LTQ Orbitrap XL, Thermo Scientfic; HP-1100 HPLC, Agilent)
• INSTRUMENT_TYPE The type of the mass spectrometry instrumentation (mandatory, e.g. LC-ESI-ITFT)
• IONIZATION The kind of ionization (mandatory, e.g. ESI) (mandatory, e.g. ESI)
• Collision energy Collision energy for fragmentation (mandatory, e.g. 15eV)
• Adduct The parent ion species (mandatory, e.g. [M-H]-)
• MS/MS Acquisition mode The acquisition mode (mandatory, e.g. DDA)
• Ionization mode The ionization mode (mandatory, e.g. Negative)
• Confidence The confidence of compound identification (mandatory, Pure standard or Predicted)
• Compound class The compound class (optional, e.g. Flavonoid)
• Found in The sample origin (optional, e.g. Tomato leaf (Solanum habrochaites LA1777))
• Database links Links to compound databases (optional, e.g. KEGG C05625; ChEBI 28527)
• Authors All authors who substantially contributed to the spectrum (mandatory, e.g. Ales Svatos, Ravi Kumar Maddula, MPI for Chemical Ecology, Jena, Germany)
• Publication Relevant pblication related to the spectrum (optional, e.g. F. Rasche, A. Svatos, R.K. Maddula, C. Boettcher and S. Boecker. Computing fragmentation trees from tandem mass spectrometry data. Anal. Chem., 2011, 83, 1243-1251)

Sheet Chromatography

Sheet for properties describing the chromatography instrumentation. Composed of two columns, namely Property with the property names and Value with the value corresponding to the property. It is recommended to use MassBank tags as property names. Please find the recommended MassBank record tags here.

Sheet Mass_Spectrometry

Sheet for properties describing the chromatography instrumentation. Composed of two columns, namely Property with the property names and Value with the value corresponding to the property. It is recommended to use MassBank tags as property names. Please find the recommended MassBank record tags here.

Install

This pipeline is designed for a Linux machine. In command line clone the slightly adapted code of the RMassBank package. git clone -b branch_treutler https://github.com/MassBank/RMassBank.git Install the package in RStudio. install.packages("[path to RMassBank source]", repos = NULL, type="source") Install openbabel. sudo apt-get install openbabel

Creating a project

A project is a self-contained data set for the generation of a set of MassBank records from raw data. Projects are currently stored at /IPB/Projects/2017_005_MS-databases/mFam contributions/. The project folder name is arbitrary and usually comprises the name and affiliation of a data contributor. The project folder structure includes the following folders.

• converted to MassBank target for the compiled MassBank records
• converted to msp target for the compiled MassBank records converted to msp format
• meta data meta data spreadsheet(s in different versions)
• raw data raw data section
• raw data/converted to abf raw data in abf format
• raw data/exported as raw msp raw msp files (exported from MS-DIAL)
• raw data/raw raw data in vendor format
• script arbitrary scripts or misc. supplemental information

The recommended workflow for the raw msp files is as follows.

• Conversion of raw data files in vendor format to abf format (see Abf converter web page here
• Raw data processing using MS-DIAL (see MS-DIAL web page here
• Export of measured MS/MS spectra from MS-DIAL to raw msp format (see settings msp format and deconvoluted in the alignment export dialog)

The meta data spreadsheet is filled independently by the experimentalist starting from the template in project template in folder meta data.

Running a project

Prerequisites

• Follow the installation guidelines above
• meta data spreadsheet under path meta data in the project folder
• raw msp files under path raw data/exported as raw msp in the project folder

Run

The conversion of raw data and meta data to MassBank records is done by RMassBank with some pre- and post-processing. To run RMassBank adapt and run the following code in the R console.

source("/mnt/ifs/data/IPB/Projects/2017_005_MS-databases/mFam_code/RMassBank_mFam/CommunityToMassBank.R")

parentFolder <- "/mnt/data/IPB/Projects/2017_005_MS-databases/mFam contributions/"

ms_type <- "MS2"
accessionPrefix <- "[TODO]"
xlsxFile <- "[TODO]"
takeRecordedNames <- TRUE
applyIntensityThreshold <- FALSE

preprocessContributorToMassBankWorkflow(folder, accessionPrefix, xlsxFile, ms_type, takeRecordedNames)
runContributorToMassBankWorkflow(folder, applyIntensityThreshold, reprocess = FALSE)

The console will print the progress of the RMassBank workflow. The MassBank records are generated in folder path converted to MassBank in the project folder. For an example of an MassBank record see here.

There are some utility function for the management of projects as follows.

source("/mnt/ifs/data/IPB/Projects/2017_005_MS-databases/mFam_code/RMassBank_mFam/CommunityToMassBank.R")

parentFolder <- "/mnt/data/IPB/Projects/2017_005_MS-databases/mFam contributions/"
aggregationFolder <- paste(parentFolder, "mFam Aggregation", sep = "")

## if a RMassBank run failes, this command cleans the project folder to the original state
#cleanContribtorDirectoryForReprocessing(folder)

## utility function for the correction of common record format problems
#correctRecords(folder)

## functions for the aggregation of spectra from multiple projects - either for all projects or a specified set of projects
#aggregateSpectra_all(parentFolder, aggregationFolder)
#aggregateSpectra_ready(processedFolders, aggregationFolder, tag = "Validated")

Test in Docker

Start the container with docker run -v $PWD/data:/data -w /data -it --rm sneumann/msp2massbank bash

and run the example data:

mkdir /tmp/msp2massbank/
cp -avx /data/* /tmp/msp2massbank/

/usr/local/bin/msp2massbank.r "XY" test.xlsx

or in one line:

docker run -v $PWD/data:/data -w /data -it --rm sneumann/msp2massbank bash -c 'mkdir /tmp/msp2massbank/ ; cp -avx /data/* /tmp/msp2massbank/ ; /usr/local/bin/msp2massbank.r "XY" test.xlsx'