Hi Abla,
At present, we recommend that people use ProteoWizard to first deconvolute the overlapped DIA data using MSConvertGUI. There is an option for processing overlapped DIA data that results in separate MS/MS spectra with no overlapping isolation. Once you have the deconvoluted mzML, you can use the "Import" button in the Skyline Edit Isolation Scheme form to get the isolation scheme for the deconvoluted DIA, which will have spectra with an isolation range 1/2 as wide as your instrument collected.

Sorry, we have gotten a little behind in Skyline, and the demultiplexing algorithm it uses by default is much slower than the one now implemented in MSConvertGUI, which explains why it took 2 hours. If you do conversion first (which you can also do from the command line with msconvert.exe), then the files should import into Skyline nearly as quickly as your DDA data.

Thanks for posting to the Skyline support board and good luck with your DIA data processing in Skyline.

--Brendan

Great, thanks a lot Brendan. will try it out
Abla

Hi Brendan,
Can you please clarify what you mean by " .. isolation range 1/2 as wide as your instrument collected". Did you mean that, say if I collected the data at an isolation window of 20mz with overlap; after I deconvolute in Proteozwizard, should I use a window of 10mz when importing data in Skyline and unselect overlap?
Thank you;
-Abla

Yes, exactly. You definitely want to unselect "overlap", because the converted data requires no further demultiplexing than you have already applied during the conversion. As for the isolation scheme itself, I highly recommend using the "Import" button to import the resulting isolation scheme from one of your files. This will give you a chance to better understand the result of demultiplexing, but yes it will be 10 m/z ranges, if you started with 20 m/z overlapped.

--Brendan

it worked! many thanks
Abla

Is there any reason why, when importing DIA data with overlap after deconvolution, I have 3% less precursors selected form my target list compared to when importing without deconvolution, even though I have the same number of targets?
thanks
Abla

Have a close look at your isolation ranges. I suspect it has something to do with 10 m/z on one edge not being included in the deconvoluted results because it is only getting measured every 2 cycles. e.g. 1000-1010.

If you look very closely at the deconvoluted versus not deconvoluted, I think you will notice a difference in the range covered, and that this will help you make sense of what is going on.

--Brendan

Thanks Brendan,
I looked at the mass range and yes, without deconvolution, a 10mz is added (at either edge actually, in alternate cycles) while this does not seem to be the case with deconvolution, which could explain why deconvolution gives less peptides.

However, I also noticed that, in the file where deconvolution was done, the number of peptides after export is even lower than before export while without deconvolution the number of peptides before and after export is the same.
I am attaching a sheet with peptides count before and after export for both files (with and without deconvolution).
thanks
Abla

Hi Abla,
I am not understanding what you mean by "before export" and "after export". What form of export are you referring to? File > Export > Report? How do you get the "before" list?

Sorry, doesn't make a lot of sense. Did you set up your document with the exact isolation list you got from using the Edit Isolation Scheme - Import button on a deconvoluted file? This should restrict your target set to only things that can be extracted from those files.

Note that I consider it a bug that Skyline itself is allowing you to extract an extra 10 m/z on either side of the consistently repeated isolation scheme, i.e. m/z range that is only repeated every 2 cycles. But, again, this is an older implementation that we will remove rather than fix.

Hopefully, we can get you clear on how to use the MSConvertGUI implementation as it is supported by Skyline, and keep you from wanting to use the older much slower and apparently buggy solution that was the original implementation.

--Brendan

Thanks for your response Brendan. By export I mean export report and "before list" is the target list.
I will repeat the work to double if I made any mistake somewhere during the process.
In the meantime, thanks for all your help as always.
Abla.