Retention Time Alignment - small molecule

Retention Time Alignment - small molecule molly hopper  2023-03-27

Hello! We're working with small molecules and while our chromatography methods are fairly stable, we will sometimes have small shifts in RT across batches that require us to manually integrate each replicate as the linked integration doesn't adequately capture the peaks. We aren't looking to predict RTs such as with iRT, but rather to align each peak within a batch. We tried view > Retention Times > alignment but because we don't have a reference library built with our own compounds we don't have a library to enter. Is there another feature or method we're missing? Our core is making the switch from vendor software to skyline, but are missing the alignment option. We've only encountered errors when following the wiki to download proteome wizard to use BiblioSpec (windows) and can't figure out how else to make a custom library. Any input is appreciated - thanks!

Brian Pratt responded:  2023-03-27

If you have stable chromotography you can declare the expected RT in your transition list.

But perhaps we can explore the library option. What have you tried, and what kind of errors are you seeing? Screen shots are helpful.

Thanks for using the Skyline support board!

Brian Pratt

molly hopper responded:  2023-03-28

Thank you Brian,
We do specify the RT in the transition list - however this doesn't perform the RT alignment that we are used to for example in compound discoverer and mass hunter. It's a part of our standard workflow and we'd like to be able to do that with the switch to skyline to ensure peak integration fully captures each injection When we try to do RT alignment options are greyed out unless we use the library.

If we try to install BiblioSpec to be able to build a library we get an error message which I've included a screenshot of. We've tried installing through Microsoft visual studio in case it was a compiler issue and get the same error.

Nick Shulman responded:  2023-03-28
BiblioSpec does not get installed when you install ProteoWizard, but it is included as part of Skyline.

The usual way to build a library in Skyline is by going to:
Settings > Peptide Settings > Library
and pushing "Build" button
or by doing:
File > Import > Peptide Search

Alternatively, if you want to run BiblioSpec from the commandline, the executables "BlibBuild.exe" and "BlibFilter.exe" are both available in the same folder as Skyline.exe.

The usual installation of Skyline installs into a folder somewhere deep inside of the "AppData" folder in your home directory. If you would like Skyline to be installed into a folder that you choose so that BlibBuild.exe is easier to find then you might want to use the Adminstrator Install of Skyline:

Most people who use BiblioSpec with Skyline just use either the "Build" button on "Settings > Peptide Settings > Library" or use the menu item "File > Import > Peptide Search". It is rare that you might want to run BlibBuild.exe and BlibFilter.exe from the commandline yourself.
-- Nick
Nick Shulman responded:  2023-03-28
Oops, I missed the part about that you were using small molecules.
For small molecules, "Import > Peptide Search" and "Settings > Peptide Settings > Library > Build" will not work.

One way to create a spectral library for small molecules would be, after you have extracted chromatograms, to use the Skyline menu item:
File > Export > Library
That will create a .blib file with "spectra" in it where the "intensities" in the spectra are based on the chromatogram peak areas.

Exporting a spectral library in that way would be useful if you wanted to create a library that would be used for improving peak picking by telling Skyline what to expect the relative transition peak areas to be.

I don't think that a spectral library exported in that way would be helpful for what you want to do in terms of retention time alignment.

We might be able to give you better answers if you send us your Skyline document.
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 and spectral libraries.

If that .zip file is less than 50MB you can attach it to this support request. You can upload larger files here:
-- Nick
molly hopper responded:  2023-03-28
Thanks Nick,
We stumbled on the wiki for the BiblioSpec Spectral Library when searching the site for how to make a library and follwed the "download and build" page not realizing that it was already part of skyline. It's good to know that it's really only for proteins anyways.

I've attached a very small skyline document for a pilot project, so it's not completely representative of the number of molecules we're normally looking for, or the complexity of the sample but it gives an idea of how we're setting things up. I've already done some manual integration on these so alignment probably wouldn't change much. But we'd like to be able to align all the peaks from replicates together as one of the first steps before we continue with the analysis.
Nick Shulman responded:  2023-03-28
Thank you for sending us that file.

Can you give me an example of a molecule and replicate where Skyline is picking the wrong peak? I might be able to give you advice about what to do about it.
I see that you are monitoring the precursor transition of light and heavy versions of several molecules.
I see that at "Settings > Molecule Settings > Labels" the label type "heavy" is marked as an internal standard.
When there is an internal standard, Skyline only looks at the internal standard chromatograms when choosing peaks.

If you happen to want Skyline to look at both the light and heavy chromatograms when choosing peaks, you could either check the checkbox next to both light and heavy in "Settings > Molecule Settings > Labels" (or uncheck the checkbox next to them both).
-- Nick
molly hopper responded:  2023-03-28
Because that file has already been manually corrected I can't give a good example on that set, but we've seen it select the isoleucine instead of leucine peaks before in cases where the RT shifted from the expected. We would like for it to look at the heavy compounds (ILE and LEU are different masses that way). It's hard for me to give a specific example because we don't want just one thing "fixed" there isn't really an example of it being completely off, except that we are used to doing peak alignment as a part of the workflow in other programs and wanted to continue that here because we know it has been benifical in the past. Is there a support option to have a phone call or video meeting so I can describe a bit better what we've done in the past and what we're hoping to accomplish? Thanks!
Nick Shulman responded:  2023-03-28
I will send you an email directly.

The retention time alignment features in Skyline might not be what you are looking for. When Skyline does retention time alignment, what it is doing is changing the way that chromatograms are presented to the user. Retention time alignment does not affect the way that Skyline sees the data. In order to get Skyline to pick peaks better, you need to give Skyline a better predicted retention time. I think iRT is something that people have used effectively with small molecules. I cannot think of any other way to deal with retention time shifts between replicates, but maybe someone else on this support board will have ideas.
-- Nick
Chris Ashwood responded:  2023-04-07
Hi Molly,

I think iRT might be a good use case for what you're doing. There are some manual steps involved (defining easily selected peaks by skyline to act as retention time alignment markers) but it skips the need to build a reference library. The only caveat is that you need five or more analytes in your run that Skyline has no trouble with picking correctly. Those are then used to normalize the retention times for the remaining analytes.

I'm not part of the Skyline team but have helped write an alpha version tutorial for iRT (Attached pdf and Skyline assay found at While the isoleucine/leucine problem is not in this dataset, I am optimistic that it will help reduce the amount of manual peak picking you need to do.

The spectral library approach might help if you have diagnostic/specific product ions for each of your analytes, but for quick testing, I think iRT is your best bet.

Chris Ashwood