Post rmb-mix QC
As suggested by @todor.kondic we should keep code ideas as issues and not (just) emails. Here are code snippets sent to @anjana.elapavalore last night to start post-rmb-mix QC discussions, based off existing ReSOLUTION functions.
Comments:
- more sophisticated plotting needed (e.g. inspired by Shinyscreen and not ancient ReSOLUTION functions)
- work towards some automated post-RMB QC criteria, e.g.
- min. precursor intensity (some were below ~2000)
- precursor intensity vs MS/MS base peak intensity (prec should not be too much smaller than the MS/MS base peak)
- min. base peak of MS/MS intensity (some were below ~8000)
@todor.kondic yes Shinyscreen is a first step but RMassBank does recal, eliminates noise etc etc so should recognise if interferences are present (and remove them) that may not be obvious in prescreening, so we will have to do post-RMB QC as well. Michele had also set up a routine in RMassBank that we may want to try, my sketch up last night was just a start. @adelene.lai may remember where Michele has notes, since he presented it at Dagstuhl and she took notes ...
We may also have slight settings discrepancies that may need to be investigated further, we can interrogate once we have a summary over all the records.