Chrom Lib "creates" new transitions

Chrom Lib "creates" new transitions tvaisar  2017-10-18 06:17:12
I found what I think is a critical issue - I tried to recreate the Skyline document from a Panorama clib and uploaded the clib from Panorama into Skyline - in the list of peptides and transitions, a new transition/fragment ion appeared for a peptide which was not there in the Skyline document used to create the Panorama clib. See the attached powerpoint.
It is a b-ion for which m/z is the same as a y-ion which was in the original. The correct y-ion did not show up in the peptide/transition list.

This actually happened for several peptides - this is just an example.

vsharma responded:  2017-10-18 11:18:19
Hi Tomas,

How are you re-creating the Skyline document from the chromatogram library (.clib file)? Can you send me the .clib file? Panorama should not add any transitions to a library that don't already exist in the source Skyline document.

tvaisar responded:  2017-10-18 11:26:27
I created the new document by uploading FASTA file with sequences of all the proteins in the clib, and then added the clib to the document.

.clib attached (there is also clc file with the same name so I attach it too.

vsharma responded:  2017-10-18 13:35:01
Hi Tomas,

Thanks for uploading the library file. I opened the file in SQLite Expert Personal, and verified that it has the correct fragment ions (y9, y10, y11, y15++) for the precursor K.FSVPAGIVIPSFQALTAR.F. It does not have the b8 ion from your screenshot.

I can't open your .clib file in Skyline. It is missing some information that is now required by Skyline in chromatogram libraries. It appears that the current community edition of LabKey Server 17.2 (that you are using) does not include the fixes that were added to support the new chromatogram library format. I will email Josh separately about this, and include you.

Could it be that the library you are adding to your Skyline document is not the chromatogram library built on your LabKey Server, but another spectrum library? The default view for chromatogram libraries, in the Library Match tab, are chromatogram peaks rather than the stick figures that are used for MS/MS spectra from spectrum libraries. It can be changed by right-clicking on the "Library Match" tab and un-checking "Show Chromatograms". Did you do that? Because your screenshot has the MS/MS spectrum style stick figure. If you are adding the .clib that you sent me then your "Library Match" tab, for the precursor K.FSVPAGIVIPSFQALTAR.F, should be blank with the message: "Failure loading spectrum. Library may be corrupted".

While you wait for the next update to the LabKey Server community edition, you can build a library in your project on PanoramaWeb.

tvaisar responded:  2017-10-18 13:49:06
Thanks. First - I repeated creating a new Skyline document and replicated the problem - again the b8 ion shows up instead of y15++.
I am positive that the chromatogram library is the one generated through the Labkey server (I do not have any other chromatogram library). I am adding the clib I sent you.
Yes - the default view was that of "chromatogram" rather than "MS/MS spectrum". It did not show chromatograms, but just straight horizontal lines so I switched it to "spectrum" view.
This is the Labkey version I have installed - Labkey17.2-52553 (community).
This is the build for the Targeted Module - 52553.21, build time - Jul 29, 2017 1:35:22 AM
 Let me know what Josh says. Perhaps there is newer build of the 17.2 available - I upgraded about 3-4 weeks ago.

vsharma responded:  2017-10-18 14:46:15

Here is the library I built on PanoramaWeb. Can you please try this one and let me know if you still see the unexpected b-ions in your document with this library?

tvaisar responded:  2017-10-18 15:08:39
I must disappoint you - the b8 ion shows up even with your chrom library. Attached is the FASTA file I use to create the Skyline document from scratch. Please try to see if there is something funky with my Skyline set up.
It also shows up in the Library Match - in legend it shows b8=771.4 (rank 4), y15++=770.9, but in the Targets list it ends up showing only b8.

Nick Shulman responded:  2017-10-18 15:42:43
I think the difference in behavior that you are both seeing depends on the transition settings.

One important setting is the "Ion match tolerance":
Settings > Transition Settings > Library

If that tolerance is sufficiently large, then the 450.76 transition is able to match either y8++ (450.7555) or b8++ (451.2369).

The other important setting is the "Ion types" on:
Settings > Transition Settings > Filter

It turns out that the order of ion types specified there makes a difference, so if you have it set to "b, y", Skyline will preferentially decide that an ion in the library is "b" instead of "y".

I would recommend that you set the Ion match tolerance to a smaller number, such as .01.
vsharma responded:  2017-10-18 15:54:40
"Ion match tolerance" is probably what the difference is. Tomas, the document you uploaded for me earlier has a value of 0.4 for this setting. I assumed that it is the same in your document. This is enough to distinguish between y15++ and b8 for this peptide. If I set it higher, I see the b8 ion annotated in the library spectrum.

Thanks, Nick!

tvaisar responded:  2017-10-18 16:12:55
Vagisha and Nick,

Thanks a lot for the responses - indeed these two parameters make the difference - even if you leave the ion match tolerance large changing the ion type from "b,y" to "y,b" results in y15++ showing up instead of b8+. And of course narrowing down the ion match tolerance also gives preference to y15++.
So it means that when you are using chromatogram library the type of the ion (i.e. y15++) is NOT extracted by Skyline, but Skyline is matching the theoretical ion m/z to the m/z indicated in the library file regardless of the ion type. Good to know. I was expecting that all this info is taken up from the library by Skyline.