Skip to content
vladpetyuk edited this page Sep 30, 2014 · 16 revisions

Notes about R/Bioc for proteomics.

Ideas:

  • Label-free quantification
  • MS QC support (integration with QuaMeter from David Tabb). Related: qcmetrics for a general framework to define and bundle QC items and proteoQC.
  • Better fragmentation pattern prediction for spectra annotation - what is, if any, already available? What do we mean by that? Just masses of b & y, a & x and c & z fragment ions? Or also their relative intensities based on the fragmentation method (CID, HCD, ETD, ECD, IRMPD, ...)? The former is trivial, the latter is challenging. See MSnbase:::calculateFragments for the former. - I didn't know MSnbase had this. I was mostly looking for the former, but extended to more esoteric ions and with support for modifications. The latter would be nice, but not something I have need for atm.
  • Add handling modifications
  • Add deisotoping capability - what is, if any, already available? Need to look up if this capability is available in other packaged. However it is quite unlikely. The good old algorithm for this kind of task is THRASH. It has been implemented in this GUI only standalone software Decon2LS
  • Handling of top-down data
Clone this wiki locally