The quantification of MS2 ions in SWATH analysis sunbergsoon  2019-08-02


Recently we compared the DIA quantification results from PeakView and Skyline, we found that the results are quite different.

We analyzed protein extracts from different developmental stages of a fly species with two replicates for each stage. We used the same spectral library to do searching in both the software. In PeakView, we exported the results using default settings and q-value cutoff=0.05. However in Skyline, we exported the results with q-value cutoff=0.05 after training a mProphet model. In skyline we used the normalized area to do quantification where the normalized method is "none".

We compared the results by computing ratio of common peptides in common proteins yielded by the two software.The comparison result is shown in attached plot. Could you please give some suggestions on explanation about the results?



Brendan MacLean responded:  2019-08-02

You should not be expecting these tools to produce the same numbers exactly, unless you are using some form of calibration. Instead, you should be comparing by linear regression and finding that they two tools have very high correlation. This will mean that a 2-fold change in abundance will produce a 2-fold change in peak area in both tools.

This was found to be the case in Navarro, Nat. Biotech 2016. No effort has been made on either side to produce matching raw peak area numbers.

Hope this help clarify what you should expect. I recommend reading Navarro, Nat. Biotech 2016 for insight on how to make the comparison.