[FIX] Correction of the observed mass if isotope error is given

[sophiamersmann]

If an isotope error is given (e.g. when using the search engine MSGF+), the calulated mass is corrected by subtracting (isotopeError * protonMass) / charge.

[jpfeuffer]

Related to #3466

[timosachsenberg]

C13C12_MASSDIFF_U usually fits slightly better to the isotopic mass distances

[timosachsenberg]

thanks! did you check if different search engines: e.g., Comet, MSGFPlus and our SimpleSearchEngine write use the same MetaValue key? and also numbering scheme (e.g., one could use 1, the other -1)
edit:
hmm did not find a meta value for comet. We might need to check if this is written out.

[timosachsenberg]

Hmm did not find the meta value for comet. Can check tomorrow, maybe we don’t parse it?

[timosachsenberg]

we also use "isotope_error" error and use 0 for the monoisotopic peak and 1 if the experimental precursor annotation points to the first isotopic peak

[timosachsenberg]

I am not 100% sure that the caluclation is correct.
calc_mass as in this code does not need any isotopic correction as it is calculated from the theoretical composition. The actual istopic misassignment error is on the experimental precursor mass.

[sophiamersmann]

You are absolutely right about calc_mass / exp_mass. I updated the code - exp_mass is now corrected rather than calc_mass.

[sophiamersmann]

Here are two histograms that show the feature DM (delta mass: difference between the experimental and theoretical mass) before and after the correction for MSGF+ search results: https://sophiamersmann.github.io///2018/07/10/isotope-error-correction/

[sophiamersmann]

I will need to check if different search engines use the same MetaValue key and numbering scheme.

[timosachsenberg]

stof uses the current locale and might thus not work for files produced on a system with a different locale.
use the toString().toFloat() instead (if toFloat() doesn't work)