by
lmh » Tue Jul 28, 2009 12:30 pm
(1) so far as I'm aware, Chemstation only allows subtraction of chromatograms, not addition.
(2) I wouldn't recommend it anyway, as small variations in retention time (leading to not-quite-aligned peaks) can lead to very strange effects at the start and end of a subtracted/added/averaged peak.
(3) you could, if you really really want to try this, export your chromatogram from Chemstation (in file menu), open in Excel, and do the averaging there. If you're working in LC-MS, I'd suggest using a package such as xcms to look at the data exported into .cdf, and do the averaging there, but this is rather specialised and hard work.
(4) but why not just integrated the peaks as normal, and average the peak areas, rather than averaging the raw data?
I suspect you are doing this because the chromatogram is so noisy that integration is difficult. If this is the case, I would generally try the following steps:
(a) can I increase signal:noise ratio, e.g. by using more sample, choosing my wavelength carefully, using a different detector, etc. etc.
(b) if not, then can I use smoothing? Some people treat smoothing as "manipulating the raw data", but since the data are already smoothed electronically (what do you think the detector's time constant or time-filter is doing??), this is an artificial argument. It's certainly safer to use a controlled numerical filter on the data than to average between two chromatograms of unknown comparability.
