search

Rappsilber-Laboratory / XiSearch

XiSearch
Apache License 2.0
9 stars 7 forks source link

xiSearch: (Isotope cluster) annotation sometimes fails due to M-1 (M-2) peaks #33

Closed lars-kolbowski closed 5 years ago

lars-kolbowski commented 6 years ago

UVPD fragmentation spectra seem to show M-1 (M-2) peaks which leads to wrong cluster annotation. just as a reminder @lutzfischer. You suggested adjusting a filter you had in place for incompletely labeled cross-linkers to ignore low intense leading peaks.

lutzfischer commented 6 years ago

the latest dev-version has some settings to help with that. You can define cross-linker losses and instruct xi to also look for the non-neutral loss-version of a fragment in the spectra explicitly. Meaning if it matches a lossy fragment of some kind to an isotope-cluster (not single peaks - as I think that would just increase the random chance for matching peaks to much) it will look if at the m/z of the non-lossy version of the fragment it can find an isotope-cluster with the correct charge state. If it finds one. it will annotate it with the non-lossy version of the fragment. This is currently only in a dev version - so keep it open until it becomes a public version.