Metabolomics Society Forum

Scientific themes => Compound identification => Topic started by: phylloxera on March 26, 2021, 02:12:02 PM

Title: level 1 identification (best and/or fast practices)
Post by: phylloxera on March 26, 2021, 02:12:02 PM
Hi,

Still trying to figure out how to ask good questions in this informatic space...  :)

Does anyone use nist mspepsearch (https://chemdata.nist.gov/mass-spc/ftp/download/peptide_library/software/current_releases/MSPepSearch) (batch mode) for level 1 feature identification in a non-targeted workflow? How about COSS (https://biblio.ugent.be/publication/8671256/file/8671294.pdf)? I was initially wanting to use the former, but can't seem to find a good tutorial/example for the command-line.
What do y'all use for this task? I'd like to use something command-line (something linux/dockerized would be even better)! I am also pretty amenable to R. my output features are in individual .msp and .mat files output from RamClustR::do.find.main. As I understand it, this wraps and applies some heuristics to CAMERA for identifying the precursor ion and adduct. Our in-house spectral library is in a single msp file. But I'd like to use a larger reference to see if our in-house spectra will be higher scoring matches to 'knowns' within the same compound, if that makes sense. I have used MS-DIAL's functionality as part of their larger workflow, but don't think there is a way to use that tool just for spectral search apart from peak-detection, deconvolution and alignment.

Thanks!
Aaron