Correction of high-resolution MS data

Thanks for this work and for having released the code of elemcor!

We have used Accucor (Su et al., 2017) to correct our HRMS data, and we noted that the correction matrix was wrong in some situations.

When reading the Elemcor publication, we realized that you also noted that calculation of exact mass differences for isotopes of each individual element was not correct, and you have implemented an elegant formalism to solve this issue (thanks again for the open source code!). However, it seems that calculation of exact mass difference is realized for individual elements, but not when combining elements. Consequently, information on the total isotopic mass defects of each isotopic species of the molecule – and thus information on the ability of the MS analyzer to distinguish them – is lost when elements are combined, and the correction matrix calculated for non-tracer elements is not correct.

We have developed a major update of IsoCor (DOI: 10.1093/bioinformatics/btz209) to solve the issue of Accucor, with the algorithm detailed in the publication. We believe that checking the mass difference of combined elements in Elemcor would required only minor changes. Exact mass differences are currently calculated by Elemcor using combinatorial probabilities (i.e. binomial or multinomial law) for individual elements to determine which combination of its isotopes can be resolved from the tracer isotopologues, but Elemcor does not account for the combination of the different elements. This could be done by calculating the exact mass of each isotopic species (i.e. after combination of isotopes of all elements), and checking afterwards which species are resolved and should thus be considered in the correction matrix.

We would be happy to have your feedback on IsoCor 2 and to discuss more generally about correction of MS data.

Thanks, @pierremillard & @bdelepine

4dsoftware / elemcor

Correction of high-resolution MS data #2