Is the ppm parameter in Centwave both sides or single side?

[amnahsiddiqa]

Hi Johannes,

From the tutorial on analysis of XCMS (https://bioconductor.org/packages/devel/bioc/vignettes/xcms/inst/doc/xcms.html#3_Initial_data_inspection- Please see attached screenshot for exact text) my impression is that the ppm parameter in centWave is both sides from mean; Could you please verify that for me ; So to say if 5 ppm is user defined ppm - the total difference in a max and min centroids in an ROI could be 10 ppm?

Thanks, Amnah

[stanstrup]

Your interpretation matches my memory and what I understood from the code checking it now.

[jorainer]

Looking a the code (not that easy because it's C code and also not mine):

for each new scan:

• calculate the absolute difference between the mean m/z of the ROI and the m/z of the current scan
• if that absolute difference is < ppm * m/z of current scan:
  • merge the current scan into the ROI and update the mean m/z of the ROI
• other wise start a new ROI

so, I can confirm what Jan wrote, due to the abs it can be that the m/z difference between individual scans is > ppm. It is hard to say if the max is exactly 2 * ppm, because the comparison is always done against the mean m/z of all data points within the ROI.

[amnahsiddiqa]

Thank you very much both of you. This was very helpful and I am clear on this now.