Problems with the EIC images and tsv report file

September 19, 2013, 02:17:05 PM

After running the following code:

>reporttab<-diffreport(xset3, "prestudy", "poststudy", file="C:/Results/pre_vs_post/20130917", 10, metlin=0.15, h=480, w=640)

I got 10 EIC images and a tsv report file. When I looked at and compared the images with the detailed list in the tsv file, I got several problems that need to be clarified by XCMS experts here.

According to one of the EIC images, it is an extracted ion chromatogram: 433.12-433.33 m/z. However, in the tsv file, the corresponding ion's mzmed, mzmin and mzmax are 433.245461, 433.243672 and 433.272257. I wanna know why the m/z range in the EIC is much broader than the one determined by mzmax and mzmin. Is the ideal m/z range 0.3 (double of the value of argument metlin)?

A similar concern is that the retention time range (highlighted zone) used for integration appears to not match with the range determined by the corresponding ion's rtmax and rtmin.

Is there any way to find out the curve details on the EIC images (just like the legend in the retention time deviation plot, which is generated by running retention time correction)? In other words, is there any way to see which sample is contributing to a specific curve of interest on the EIC image?

On some EIC images, the highlighted area covers several obvious peaks. What is the reason for this? To my understanding, the highlighted area should only have one peak. If so, how to improve the peak picking process?

In addition, what is the default value for argument eicwidth in method diffreport?

Your early response will be highly appreciated.

Best wishes,

Jeff

Re: Problems with the EIC images and tsv report file

Reply #1 – September 19, 2013, 03:43:42 PM

Jeff,

Quote

According to one of the EIC images, it is an extracted ion chromatogram: 433.12-433.33 m/z. However, in the tsv file, the corresponding ion's mzmed, mzmin and mzmax are 433.245461, 433.243672 and 433.272257. I wanna know why the m/z range in the EIC is much broader than the one determined by mzmax and mzmin. Is the ideal m/z range 0.3 (double of the value of argument metlin)?

This is due to the way in which the EIC's are generated. They are made using a binned matrix of the LC-MS data file. Consequently, the actual mzmin-mzmax range might be different. You could manually make the EIC's using rawEIC().

Quote

A similar concern is that the retention time range (highlighted zone) used for integration appears to not match with the range determined by the corresponding ion's rtmax and rtmin.

This I don't understand. It should be the same range. How much is it out by?

Quote

Is there any way to find out the curve details on the EIC images (just like the legend in the retention time deviation plot, which is generated by running retention time correction)? In other words, is there any way to see which sample is contributing to a specific curve of interest on the EIC image?

Do you mean from the fold difference? The fold difference is taken as the median of each class and then class1/class2.

Quote

In addition, what is the default value for argument eicwidth in method diffreport?

The default values can be found from looking at the help/manpage again just use

Code: [Select]

help("diffreport") 
## or a tad quicker
?diffreport

Hope it helps, let me know how you get on.

Cheers,

Paul