Integration fails in MS-1 when multiple scan ranges are used

support
Integration fails in MS-1 when multiple scan ranges are used v delcourt  2018-11-09
 

Hi Brendan,

I've been doing some experiments with a splitted MS1 (1 scan for 200-405 and 1 scan for 400-1300 and then DIA) which are separated in scan events in the method editor. Skyline does not seem to integrate it properly. Indeed, it looks like it's trying to look for target ion in the wrong scan range (see capture below) and results in abnormal integration.

Maybe there's something i'm missing. I've been reimporting my DIA scheme, trying to play with MS1 integration parameters but couldn't find the solution.
I guess this issue could be faced in case of BoxCar analysis as well as this is kind-of the same approach (in a non-multiplexed fashion).

Best regards,
Vivian

 
 
Tobi responded:  2018-11-15

Dear all,

hope it is ok if i make a comment since we have a similar issue in our lab, just with 2 AIF elements of different ranges in one measurement. (Same issue as described above, just on MS2 level)

You can see on the screenshot (and my attached file) that every second chromatogram point has the intensity of close to 0, since for those points skyline integrates the 770 m/z peak from the 200-405 scan.

Is it possible to solve this issue, for the variants MS1 and MS2 directly in skyline? Do you know a workaround (splitting into 2 seperate files before integration)?

Best regards,
tobi

 
Nick Shulman responded:  2018-11-15
If Skyline sees an MS1 spectrum whose m/z range is less than 200 units, then Skyline considers that to be a "SIM" scan, and Skyline will only try to extract a chromatogram point for the precursors that are in the range.
If the range of the MS1 spectrum is more than 200 units, then Skyline thinks that it is a full MS1 scan, and Skyline expects to find every precursor in that scan, and anything outside of the range that was scanned will get a zero in the chromatogram.

I am not sure whether the best way for us to fix this in Skyline would be to give you more control over what Skyline considers to be a "SIM scan", or whether we should make Skyline smarter about ignoring MS1 spectra which do not include the precursor.
-- Nick
 
v delcourt responded:  2018-11-15
Hi Nick,

Thanks for your answer !
I thought it might be convenient for Skyline to treat MS1 as it does for DIA-MS2 scans, which means that there could be a MS-1 acquisition scheme which could be imported in the same way it's done for DIA. If done so, Skyline would integrate each precursor according to their specific scan ranges. This strategy could be used for BoxCar analysis (even if it's effectively) treated as tSIM scans if I understood properly.

One other option would be to parse scan headers (ms level, isolation width+scheme and acquisition range) and then integrate if target (precursor or fragment) ion is included in scan header but I don't know if it's possible...
I believe this would solve both issues for MS1 and MS2 as (Tobi described).

In my opinion, your second suggestiont ("make Skyline smarter") would be the best but I don't know what are the implications on the software's structure

Best regards,
Vivian
 
matt.chambers42 responded:  2018-11-16
Nick, what is the reason for not checking to see whether the lower and upper limits of the scanWindow are available and using those to decide whether to use the scan for a particular precursor? You can fall back on the heuristic if the scanWindow isn't available, which should only be for very old data.