Waters HDMSE Raw Data Visualization brad williams  2020-04-13
 

I have been testing the latest Skyline builds (20.1) and Skyline Daily (20.1.1.83) and they both do not allow the raw data to be visualized for MS1 precursors. The raw data plot shows up blank and the MS/MS information is there. Has this been reported before? I have attached a screen capture to show an example. The datafile type is HDMSE.

 
 
Nick Shulman responded:  2020-04-13
Can you send us your Skyline document and one of your Waters .raw files?
In Skyline you can use the menu item:
File > Share
to create a .zip file containing your Skyline document and supporting files including extracted chromatograms.

You should also send us your Waters raw data. I believe that means that you should zip up one of your .raw directories and send that to us.

You can upload those .zip files here:
https://skyline.ms/files.url

When Skyline extracts chromatograms, Skyline remembers the "spectrum identifier" of each spectrum. Skyline represents these spectrum identifiers as a short list of numbers separated by dots. If Skyline is unable to show you some spectra, it might mean that the spectrum identifier that Skyline saved with the chromatogram is incorrect, or for some other reason, cannot be used to find the spectrum that Skyline is trying to show you.

We will be able to figure out what is going wrong when we see your files.
-- Nick
 
brad williams responded:  2020-04-16
Hi Nick,

Thank you for the reply. I uploaded the archived skyline file and an example raw data file. The file was compressed with WinRar and will need to be uncompressed back to the .raw form. I have uploaded a screen capture of the file names that have been uploaded to the Skyline datashare.

Thank you,

Brad
 
brad williams responded:  2020-07-30
Hi Skyline,

I have been testing the HDMSE visualization of the MS1 and MS2 spectra in the lastest version of Skyline (20.1.0.155) and the IMS filtered spectra still show up blank. The Skyline_HDMSE.PNG shows an example skyline project and HDMSE data file to test this when you have a chance.

Thank you,

Brad
 
Nick Shulman responded:  2020-07-30
Brad,

Oops. I forgot to take a look at this one when you first posted this.

The file that you uploaded was
2019_12_12_K562_Res_UDMSE_100ng_01.raw
but the file that you are showing in the PowerPoint is:
2019_12_12_K562_Sens_UDMSE_100ng_02.raw

The Full Scan viewer looks fine with the file that you provided.

Can you send us that other file?
-- Nick
 
brad williams responded:  2020-07-31
Hi Nick,

No worries. The file should not matter. I have tested HDMSE data from the latest SYNAPT XS and the SYNAPT G2-Si. Both HDMSE datasets show a blank spectrum when you apply the IMS filter. The MS1 and MS2 spectra appear just fine with the filter deselected. What happens when you apply the IMS filter to the raw datafile that I provided?

Regards,

Brad
 
Nick Shulman responded:  2020-07-31
Oh, I see what you are talking about.
The spectrum ends up being blank if you push the "filter" button on the Full Scan graph and you are looking at an ion mobility scan in 2D mode.

This seems to be a problem with how the Full Scan graph is handling the new "combined ion mobility mode" where ProteoWizard gives us a single spectrum with three arrays: m/z, intensity, and ion mobility. (You could probably work around this bug by using MSConvert and making sure to leave "Combine ion mobility scans" unchecked)

We will fix this.
-- Nick
 
Brendan MacLean responded:  2020-07-31
Please, wait for the fix and don't try what Nick suggested. This leads to truly massive files (e.g. in the tens of GB) because the MSConvert formats (mzML and mz5) are so heavy on the header data for each spectrum, using 2D spectra to represent IMS data works really poorly. That is a big part of how we realized we needed to fix this, but apparently this is one case we missed in the transition.

Thanks to you and Nick for figuring this out.

--Brendan
 
brad williams responded:  2020-08-03
Hi Brendan,

Ok, I will wait for the fix. I can test in Skyline Daily once you have it fixed. Just let me know.

Thank you,

Brad
 
Brian Pratt responded:  2020-08-03
Hi Brad,

This will be fixed in the next Skyline-Daily, but for now all you have to do is turn off the Filter button in the Full Scan window. and all should be well.

For the more modern 3-array IMS data representation (mz, intensity, mobility) it has little effect anyway - in the old less efficient 2-array (mz, intensity) representation it was intended to allow you to skip over scans with ion mobility outside the range of interest, but with the new representation such scans don't really exist. So it's OK to just leave it turned off, when using the left and right buttons to explore nearby scans you will seldom be hitting anything that doesn't have data within the IM range of interest..

Best,

Brian Pratt
 
Juan C. Rojas E. responded:  2020-08-04
Dear Skyline support team,

I post this request here because of the relevance to the issue posted by Brad, sorry if it was meant somewhere else.

Background details:
- Instrument: Synapt G2-Si
- I create spectral libraries through IMS-DDA of a pooled sample prior to acquisition of my replicates; this data also hold the ion mobility of precursors for confirmation
- Replicates are measured in HDMSE with fragmentation in transfer cell; fragments and precursors will have same mobility
- There are slight differences between the ions recorded in IMS-DDA with "wideband enhancement" and those recorded in our HDMSE setup.

I am using the IMS prediction tool provided in Skyline to find the best IMS extraction window (and correct list manually if necessary) with "Use Results". Then I apply this window to reintegrate my .RAW files to clean-up interferences. At this point in the 3D visualization you can nicely observe the the extraction window in the IMS and m/z dimension.

Would it be possible that when the filter botton is ticked the MS (in the high energy scan) spectrum displayed in 2D show only the mz values that fall inside of the IMS extraction window? Furthermore, could the MS scan be saved/exported as a .mgf which woud contain a pseudo MSMS spectrum (whether of single spectrum or average across integrated peak) filtered with the IMS dimension? This with the purpose of generated a second library specific from the HDMSE data.

The only tool that I know of that can perform this sort of manual extraction for Waters data (without a blackbox algorithm going on in the background) is DriftScove. However, this is a very painfully slow and inpractical approach... For my particular case I cannot used PLGS with it pep3D algorithm since the search engine relies on indentification of unmodified peptides and I work with fractions of enriched modified peptides where the "backbone" unmodified peptides is not expected.

Summary request: Use Skyline to extract pseudo MSMS spectra filtered in the ion mobility dimension.
 
Brian Pratt responded:  2020-08-04
Hi Juan,

There are two parts to this, I think:
1) In 2D view, show only mz,intensiity values whose IM value falls within the filter
2) Use that same filtering to generate .mgf representations of that filtered 3D data collapsed into a 2D representation


For the first, of course, that's the intent of the Filter button in the Full Scan 2D view. But as you know from this thread, that was broken in the switch to the more efficient representation of IM data as mz,intensity,IM points instead of scans with an IM value and a list of mz,intensity pairs. So that will be fixed in the next Skyline-Daily.

For the second, this would probably be better posed as a request to msconvert support at http://proteowizard.sourceforge.net/support.html.

Best,

Brian Pratt
 
Brian Pratt responded:  2020-08-04
Juan -

re the second point about generating mgf, I forgot to mention that the "Copy data" right-click menu item in the Full Scan 2D view goes pretty far in this direction, as it puts that visualized (filtered, summed intensities per mz) information on the clipboard:

767.580444335938    0
767.592529296875    0
767.604614257813    26
767.61669921875    120
767.628784179688    97
767.640869140625    12
767.652954101563    56
767.6650390625    0
767.677124023438    85
767.689208984375    132
767.701293945313    103
767.71337890625    21
767.725463867188    0
767.761779785156    27
767.773864746094    14

Hopefully that's enough to get you where you want to be with just a little extra work on your part to create the headers etc.

Best,

Brian
 
Juan C. Rojas E. responded:  2020-08-05
Hi Brian,

Yes, the "copy data" gets me already very far on what I hoped! However, could you help me understand why there are multiple "m/z" and "intensity" sections.

The first section seem to have the regions specifically for the extracted ions whether precursor ions (in the example attached) or fragment ions matching to library. But then there are two sections with the full m/z range of the MS spectrum that seem to be duplicates. Ultimately I can edit them the initial sections and one of the duplicates manually or with an R script, but understanding why its split like this would help. Or this is something that would also be altered with the next release of Skyline Daily? If the latter is the case I will just wait.

Thank you for your help and time!

Juan
 
Brian Pratt responded:  2020-08-05
Hi Juan,

I had to dig around in the code to understand that output myself.

Each section represents a different subset of the m/z,intensity values in the graph, in this order:

points with m/z values in range of first transition (i.e. points in color A)
points with m/z values in range of second transition (i.e. points in color B)
etc
points not associated with any transition (i.e. points in grey)
all points (union of all the points in the preceding lists)

So it's that last one you probably want. Obviously this would be more useful if each section had a title, I'll fix that.

Best,

Brian