Chromatography Forum

Hi all,

I am attempting to check the reproducibility of GC/MS data. Qualitatively successive runs look the same but I would prefer to be able to show just how close they are in more quantitative manner. One way to do this that has been suggested is to subtract one run from another and (hopefully) show the resulting blank baseline. However I am struggling to do this in my program.

I presume that I could do it by moving the data into a spreadsheet but if I could do this using a proper prcessing program I think it would be better.

I am currently using Xcalibur 2.0.6 and the data is acquired using a Thermo TraceMS system. Any help or suggestions of other programs that might be able to do it would be appreciated.

Many thanks in advance

G

Hi George

A slightly unusual approach, any reason that you want to do it that way ?

I do not know your software, but you might be able to designate one of the chromatograms as a baseline correction for the other one - usually this is used to correct baseline drift but it might work.

Are you looking to compare MS m/z vs intesity scan by scan, or chromatographic peak by peak on the TIC ?

Peter

Chromatographic peak by peak was how I imagined trying to do it.

The reason I would like to do it this way is that it was suggested as a neat visual way to demonstrate the reproducibility. More specifically I am doing pyrolysis-GC/MS - each chromatogram has at least 150 compounds in it. If anyone can suggest another way to do this I would also be interested!

I'll have a look at baseline correction in our software - thankyou.

G

Hi George

I can see that working well as a series of slides. If you cannot get it to work, then just overlay the two traces in different colours - if you can only see one colour then the chromatograms match.

Peter

Exporting the integrated peaks to excel is probably the easiest way for one time event.

If you have to do it regularly then it's maybe worth considering one of those metabolomics software where you can monitor the abundancy of several peaks in an easier way (maybe Mzmine2, or similar ones).

George,

I would do 5 successive runs with one sample to yield 5 datasets. I would then deconvolute in AMDIS and import the .elu file into MET-IDEA. This would give me a spreadsheet which I can then copy into Excel to calculate the % variation for each of the metabolites found in the five aliquots of the same sample.

However, you have to have an Internal Standard to normalize the data with beforehand..

Thanks very much for the suggestions all. Armed with more knowledge I have now found a way to subtract whole chromatograms from each other using the background subtraction toolset.

Whilst I can't get to the actual numbers this way it does give a clear visual representation.

G