Question regarding DIA Isolation Window Scheme output for Thermo Fusion DIA m/z list

Question regarding DIA Isolation Window Scheme output for Thermo Fusion DIA m/z list dhardie  2019-10-16

Dear Skyline Team;

I have a question regarding the DIA Isolation Window Scheme output when using Skyline Daily 64-Bit the output table contains "Start and End m/z" with the ability to show "Margin" and "CE Range" and has the "optimized window placement" feature but there is no "Centre m/z" similar to the EncyclopeDIA Scheme Wizard? Older versions of Skyline and presentations show "Generate Target " toggle that would show a centered target m/z in the output list that could be exported as a csv into the Thermo Fusion MS2 targeted list. The current output is just start and end m/z. Do you suggest manually calculating the "centre mass" for the DIA acquisition as a work-around.

Also I have noticed the masses generated when using margin widths of 0.2 if using Skyline and EncyclopeDIA window generators give different values. I have seen Searle's bitbucket documents describing Skyline's settings for DIA windowing acquisition settings.

Is there a formal document explaining the windowing scheme settings MSX,Overlap, Overlap MSX, Fast Overlap Experimental and what these settings are? I see many presentations for Q-Exactives but not much for Lumos or Fusion.

Thanks for entertaining my questions


Brendan MacLean responded:  2019-10-27

Hi Darryl,
Can you please include some screenshots of what you are seeing. It is always a little hard to imagine the words into images, and where I think you should be looking may not turn out to be where you are at all. Are you talking about the CSV output of File > Export > Isolation List? And you are saying that the File > Export > Isolation List form no longer offers you the option of exporting center and width for Thermo instruments?

That would surprise me, but a few screenshots would tell a lot more.

Thanks for posting your question to the Skyline support board.


dhardie responded:  2019-10-28


Thanks for the reply. The File>Export>Isolation list is working perfect as it should...exporting the centered m/z. I was mistakenly just manually cutting and pasting from the Skyline calculation window......
Attached are the outputs from Skyline exported isolation list showing centered m/z. I have included the Skyline and EncyclopeDIA output lists for 0 margin and 0.2 margin overlapping DIA experiments. Once margin is enabled the values differ between these two programs. So I am doing something incorrectly? Do I have an incorrect setting in generating the windows?
Thanks for all the work your Skyline team does to provide such useful software.


Brendan MacLean responded:  2019-10-28

To me, it looks like Encyclopedia is getting this right and Skyline is getting it wrong, just based on the first window, which both agree on without margins:

400.4319 to 424.4428

But adding the 0.2 margin, makes EncyclopeDIA start with:

400.232 to 424.243

Which seems totally reasonable, but Skyline uses:

401.2324 to 425.6433

So what exactly makes Skyline jump by 1.0005 before subtracting 0.2 for the margin? We will have to look into that. It is definitely intuitively obvious.

The two tools are obviously very close, except for this jump by 1 peptide optimized boundary m/z unit (1.0005). Initially I am guessing that Skyline is the one making a mistake, but it may turn out that there is some important consideration that underlies this shift, which didn't get carried from Skyline to EncyclopeDIA.

Nick Shulman responded:  2019-10-28

I believe that the reason that Skyline makes your first window start at 401.2324 was that we were trying to fix some scenario where you might have asked for 20 m/z windows between 505 and 905 and because of rounding errors we might end up thinking that 21 windows were needed to fully cover that range.
In order to fix that scenario, I think we have ended up rounding up too aggressively, which is causing the behavior that you are seeing.

If you want to get Skyline to give you an isolation scheme that looks the same as EncyclopeDIA's Window Scheme Wizard, you can tell Skyline that you want your windows to go from 399.8 to 1000 instead of from 400 to 1000.

By the way, we are not sure whether the deconvolution code in MSConvert is able to handle overlapping windows with margins. That deconvolution algorithm subtracts signal that came from precursors in the other half of the scan, and I am not sure what happens to signal that might have come from the margin. Dr. Searle might have an answer as to whether that is something the deconvolution algorithm can handle.
-- Nick