Metabolomics Society Forum

Software => R => XCMS => Topic started by: Yasin on May 19, 2018, 04:54:54 PM

Title: Plotted integration-space does match actual integrated area
Post by: Yasin on May 19, 2018, 04:54:54 PM
Hello,

I ran into a little a problem and hope someone can help me understanding it.

I used the centWave algorithm of XCMS to detect peaks. That worked generally well. However one of the peaks I found was the one attached. For obvious reasons I looked into it. However, I found the integration indicated by the plot  did not match the area given as "into" even close. I extracted the chromatogram to test it and although it looked absolutely the same the area was just different by orders of magnitude. When I tried to find out the reason I saw that rtmin and rtmax confirm the plotted integration boarders. However, scmin, scpos and scmax are all given as -1 which does not really makes sense. lmin and lmax are confirming the values in mzmin and mzmax. Could somebody please explain this to me? I just cant wrap my head around it...

                            mz           mzmin            mzmax             rt               rtmin                rtmax              into                                                                       
28945   116.0705011   116.0703547   116.0706169   85.88142934   57.2363598   92.180892   1625626.385   
      intb               maxo    sn      egauss                   mu                 sigma                h                       f        dppm scale scpos    scmin   
1301337.896   7340575   137   0.0340379   335.2332267   2.950305611   7058107.727   56241   0   -1   -1   -1   
scmax     lmin   lmax      sample
 -1         218   351         5


Ps.: I also wonder how the rt is calculated here. It is neither the average between mzmin and mzmax nor at the highest intensity.

Thanks in advance to all the people who are contributing to this forum.

Regards
Yasin

[attachment deleted by admin]