• Sign In
  1. Requests

SRM - Force Skyline to integrate scans with different/multiple headers

support

View Request

view list print
SRM - Force Skyline to integrate scans with different/multiple headers v delcourt  2018-05-11 09:30
 

Hi,

I'm currently working on optimization in MRM/SRM using a quantiva. Let's say I didn't want to miss any transition and I asked skyline to produce a,b,y transition list which I imported into a quantiva.
After analysis, skyline integrates only a few ions which seemed surprising in the first place as i knew that some ions would be produced by CID but were missing from Skyline integration. After looking a little bit deeper into the data, I noticed that the Quantiva renamed my precursor groups of 10 transitions

Example : Precursor = 803.37 -> 18 transitions

Group 1 : 803.370 -> 10 transitions

Header : + c ESI SRM ms2 803.370 [105.065-105.067, 136.075-136.077, 164.070-164.072, 202.118-202.120, 207.112-207.114, 235.107-235.109, 354.180-354.182, 365.181-365.183, 382.175-382.177, 411.202-411.204]

Group 2 : 803.371 -> 8 transitions

Header : + c ESI SRM ms2 803.371 [422.202-422.204, 439.197-439.199, 569.271-569.273, 574.265-574.267, 602.260-602.262, 640.308-640.310, 671.318-671.320, 699.313-699.315]

The problem is that Skyline only takes into account the first group as the only group for integration and that most intense ions are contained within second group. (see capture)

Maybe there's some quick workaround for this but I wasn't able to find it. Playing with Peptide and Transition options didn't help so far. I also tried to trick skyline and asked to import it as DIA which wasn't successful either (as expected :( ).

Is there anything I'm doing wrong ?

Btw this is a sticky peptide so no surprise it has such a wide rt.

Best regards,
Vivian
 
 
Brian Pratt responded:  2018-05-11 10:31
Hi Vivian,

This will be easier to investigate if you can provide your Skyline document (use File>Share>Complete to create a tidy .sky.zip file) as well as the raw data file. You can upload them to http://skyline.ms/files.url, or we can arrange something else at your convenience.

Thanks for using the Skyline support board!

Brian Pratt
 
v delcourt responded:  2018-05-14 06:15
Hi Brian,

Thanks for the quick answer. Files have just been uploaded using the page specified in your post.

Best regards,

Vivian
 
Brendan MacLean responded:  2018-05-14 11:48
Hi Vivian,
Can I ask whether you exported your method or transition list with Skyline? It seems we have a workaround for the Quantiva which counts the number of transitions written for a specific compound name and changes the compound name after 10 have been written.

This would seem to imply that the Quantiva is capable of having more than 10 transitions with the same precursor m/z as long as they have different compound names. Skyline doesn't look at the compound names during import, but it does assume that different precursor m/z values (no matter how small the difference) mean different precursors.

If the Quantiva actually can't measure more than 10 transitions with any given precursor m/z value, then we will need to implement some other new special hack to support it, but you are the first to report this, and our lab has been using the Quantiva for years now. So, I am hopeful that this is a method crafted outside Skyline that didn't use the trick of changing the compound name, and so ended up creating the method with the 0.001 precursor m/z shift, which is currently incompatible with Skyline treatment of such changes.

Thanks for the work you have done to understand this issue and post it so clearly to the Skyline support board.

--Brendan
 
v delcourt responded:  2018-05-15 00:37
Hi Brendan,

Thanks for answering so quickly. Indeed, transition list has been generated by skyline but has been manipulated before importing into the quantiva. I first thought it was inconvenient to use "peptide modified sequences" to name compounds into the MRM transition list, that's why I renamed the compounds with the name of the targeted analytes.

Thus I did not notice the trick of skyline to add a number every 10 compound->transition in the csv and didn't apply this trick myself, that's probably why I am now struggling with data import.

I admit my data table manipulation is responsible of this issue. However, I don't know if it would be inconvenient to add a feature to Skyline to perform some kind of wildcard search on scan headers for import ? I don't really know the architecture of the code behind but in the early step of SRM import, make a list of "really close" selected precursors to group them would be a nice feature ? Something like "803.37*" or "803.3*" would gather scans with precursors in "803.370-803.379" or "803.3+/- width used for first quad selection" ? This would make sense since triple quads are (really)low-res instruments and separating scans by 0.001 m/z is not realistic. I don't know the point of limiting precurors->transitions to ten since TQ are now faster and faster and thus, one would want as many transitions as possible... Let me know if I can do anything to help.

Again, thank you for your quick support and for bringing such a great software to the community.

Best regards,

Vivian