Hi Sean,
Yes, in the end Skyline chooses the highest scoring peak in any given chromatograms, much like peptide search engines that provide the highest scoring peptide sequence for any MS/MS spectrum. Often they are wrong. The statistics of deciding when such a peak pick is wrong are still under development in all software that does this, and I am pretty confident that none are as far along as the programs that have done this for well over a decade in spectrum matching.
You can certainly export a report from Skyline that includes both the peak areas and idotp values, and do your statistical analysis, as you suggest, where you zero anything with an idotp below a certain value. Unfortunately, there are many possible such cut-offs people might want to use (certainly mass error as well), and I have my doubts about whether my collaborators in statistics would agree that zero is the right value to provide in this case.
But, we will certainly continue working on this, and trying to make Skyline do a much better job in the face of the absence of anything reasonable to pick. Many I have worked with would integrate noise in the RT range where the peak is expected. So, that may be what we aim for.
Thanks for posting your experience and your thoughts on possible solutions. While there is nothing automated or officially part of Skyline that does this, Skyline certainly gives you the information you would need to implement this in a statistics package like R, or probably even Excel.
--Brendan |