Hi Matt,
I knew this would be an area for confusion with DIA/SWATH. Originally, I wanted to design the isolation scheme editor to work like I think you are expecting, but after trying to implement that and work with it, we decided to go with what we have now.

That is Start and End in the isolation editor refer to the m/z range you wish to have ions extracted from, and not what the instrument actually isolated. For instance, the original SWATH paper had 26 m/z isolation windows, spanning 400 m/z to 1200 m/z with 1 m/z overlap to account for a 0.5 m/z margin on either side where the Aebersold lab measured signal drop-off. That method would be described as follows in the isolation scheme editor:

Start End Margin
400.5 425.5 0.5
425.5 450.5 0.5
450.5 475.5 0.5
475.5 500.5 0.5
...

If instead you described it as:

Start End Margin
400 426 0.5
425 451 0.5
450 476 0.5
475 501 0.5
...

which is probably closer to how the method was described to the instrument, then you would be allowing Skyline to extract chromatograms from the range where signal drop-off is observe, and I would expect you to see very spiky chromatograms for anything in that range, as Skyline would be extracting points alternating between good signal and signal from the drop-off range.

You still wouldn't necessarily see the error you are experiencing, but I am getting to that. The fact is that your isolation ranges overlap, when you may not necessarily want that, unless you believe you will get equally good signal in the overlapping region from both ranges that overlap.

The other danger with overlapping ranges that is the cause of the error you see is that all Skyline gets from your WIFF file is the "isolation target" for each isolation window. From that isolation target, Skyline must determine which isolation window in your isolation scheme the target corresponds to. If the target falls within two isolation ranges, then Skyline will show the error you have posted, which is telling you:

The isolation target 1196.66558852419 is contained by the isolation ranges 1,169.24 - 1,196.80 and 1,195.80 - 1,200.00.

The isolation target comes from your WIFF file and the ranges come from you isolation scheme. You can certainly fix this by not having overlapping isolation ranges in you isolation scheme, but you probably also want to check very carefully that your ranges are correct, since this error also seems to indicate a misunderstanding of the ranges, which should not be very close to the signal drop-off range. Though, unlike all other data formats, AB SCIEX WIFF files do not put the isolation target in the center of the isolation range. In other SWATH data I have seen the isolation targets were 5 m/z from the start of the range and 20 m/z from the end.

Hopefully this is enough information to help you fix your isolation scheme.

You certainly highlight the need for a tutorial on this stuff. Thanks for taking the time to post your issue with such informative screenshots.

--Brendan

Thank you for your response.

1. I failed to mention that I used the same method and isolation scheme in high-sensitivity mode and was able to import the data into Skyline with no problems. -- Thoughts?

2. You mentioned 'In other SWATH data I have seen the isolation targets were 5 m/z from the start of the range and 20 m/z from the end.' -- How do I set the isolation targets?

3. It would be best for me to generate an isolation scheme with no overlapping windows?

Best,
Matt

Hi Matt,

1. I don't know enough about this. Your instrument method must have generated different isolation targets for some reason in the resulting WIFF file. Obviously, if the WIFF had the same isolation targets and you used the same isolation scheme in Skyline, you would have gotten the same error.
2. I am not sure you can. I think AB SCIEX software sets them for you based on the isolation ranges you choose. But, I know that they will not be in the center of your instrument isolation ranges.
3. Probably, you do not want overlap in your Skyline isolation scheme, unless you really know what you are doing with it. The values you enter for Start, End and Margin should add up to what you used in your instrument method:

Instrument start = Skyline start - margin Instrument end = Skyline end + margin

Definitely double check that you are completely sure the ranges you used in your Skyline isolation scheme are consistent with what you used in your instrument method. When I look at the Start and End values I see in your screenshot, I see the following:

Start	End	Width	Overlap
400	426.56	26.56	0.96
425.6	453.12	27.52	1.02
452.1	479.68	27.58	2.88
476.8	506.24	29.44	1.04
505.2	532.8	27.6	1
531.8	559.36	27.56	0.96
558.4	585.92	27.52	1.02
584.9	612.48	27.58	0.98
611.5	639.04	27.54	1.04
638	665.6	27.6	1
664.6	692.16	27.56	0.96
691.2	718.72	27.52	1
717.72	745.28	27.56	1
744.28	771.84	27.56	1
770.84	798.4	27.56	1
797.4	824.96	27.56	1
823.96	851.52	27.56	1
850.52	878.08	27.56	1
877.08	904.64	27.56	1
903.64	931.2	27.56	1
930.2	957.76	27.56	1
956.76	984.32	27.56	1
983.32	1010.88	27.56	1
1009.88	1037.44	27.56	1
1036.44	1064	27.56	1
1063	1090.56	27.56	1
1089.56	1117.12	27.56	1
1116.12	1143.68	27.56	1
1142.68	1170.24	27.56	1
1169.24	1196.8	27.56	1
1195.8	1200	4.2

A few things appear strange to me about this isolation scheme:

1. The overlaps and widths become extremely consistent about 1/3 through, but are quite variable before that.
2. The final range is only 4.2 m/z.

It does look like you have 1 m/z overlap in the last 2/3, which means your 0.5 margin is probably correct, but you want to subtract the 0.5 before entering the values into Skyline.

Hope this makes things a little clearer. If you get the ranges correct, you shouldn't need to know the isolation targets. And, with AB SCIEX, it is non-trivial to figure out how the isolation target relates to the range.

Good luck. Sorry this isn't simpler, but at least you should only need to get this set up correctly once, and then you can just keep reusing that isolation scheme.

--Brendan

Got a little mixed up on my Wiki syntax. If that last post didn't look right in email, be sure to check the web site, where I have edited my response to use a table.