Hello all,
I am working on doing some untargeted metabolomics on human/rat/mouse tissues and plasma. Using Orbitrap MS cycling voltages (10,20,30).
When I attempt to use databases (ie: ones on MS-DIAL website) I get few annotated peaks, all 'w/o MS2" and most dont really make sense.
Any advice?
Is there a msp file for HMDB?
Thanks!
Hi Massspek,
HMDB is NOT a spectral library but reflects other spectral libraries. So if you want to focus on human metabolites, better to do as much annotations possible using multiple spectral libraries such as the folllowing and then sort the annotations using HMDB IDs.
1. MoNA .msp "ALL spectra":https://mona.fiehnlab.ucdavis.edu/downloads
2. GNPS .msp all spectra:https://gnps-external.ucsd.edu/gnpslibrary
3. Combine the rest from RIKEN MSDIAL server
You can combine all .msp's accessible as .text files, following instructions as here :https://www.protocols.io/view/steps-for-building-an-open-source-ei-ms-mass-spect-eq2ly33rqgx9/v1 when it was done for GC-EI-MS spectral libraries, but the idea is simple as once Hiroshi taught me this way!
Note: Please "ignore" using all 'w/o MS2" at all- and you are very right there.
Also when searching against the spectral libraries you got to ignore the RT values. Only MS/MS matches at 80-90% cosine similarity fora any further analysis or interpretation !
Hope this helps.
Thanks,
Biswa