search

WMBEdmands / compMS2Miner

metabolite identification R-package for metabolomic high-resolution LC-MS datasets.
14 stars 8 forks source link

Low resolution ddMS2 data #1

Closed dwalke04 closed 8 years ago

dwalke04 commented 8 years ago

Hi William, Excellent work. Can you use the MS2 processing tools with low resolution fragmentation spectra? We are collecting Orbitrap data for MS1, but ddMS2 is being detected in the ion trap at unit resolution.

Thanks, Doug

WMBEdmands commented 8 years ago

Thanks Doug,

Yes you can adjust the parameters for orbitrap data. You will just need to set the following arguments in the following functions to values appropriate for nominal mass MS/MS data:

# mzError ( in daltons e.g. 0.8-1) combines adjacent MS/MS spectra within the same file of different 
# collision energies.
object <- combineMS2(object, 'Ions', mzError=0.8)
#  mzError ( in daltons e.g. 0.8-1) combines MS/MS spectra between MS/MS files
object <- combineMS2(object, 'Spectra', mzError=0.8)
# if you wish to use the spectral similarity feature (recommended) also adjust the binSizeMS2 
# argument. 
# (i.e. only combining spectra with a dot product score above the specSimScore argument).
object <- combineMS2(object, 'Spectra', mzError=0.8, specSimFilter=0.8, binSizeMS2=1)
# this will prevent you unknowingly combining  spectra from co-eluting isobaric compounds with 
# different fragmentation patterns.

# substructure annotation (0.8-1 daltons).
object <- subStructure(object, 'Annotate', Frag_mzabs=0.8)

# and also some parameters you would need to change in the combinatorial metabolite identification
# functions.
# NIST msp database matching
object <- metID(object, 'matchSpectralDB', binSizeMS2=1, 
                mspFile='http://prime.psc.riken.jp/Metabolomics_Software/MS-DIAL/MSMS-MassBank-Curated-Pos.msp')
# spectral similarity network dot product bin size
object <- metID(object, 'specSimNetwork', binSizeMS2=1)

Make sure to watch this space as the package is still in development.

Will