Skyline-daily (64-bit) 21.0.9.139 (6dbba07d9) cannot load the library.tsv file from EasyPQP

support
Skyline-daily (64-bit) 21.0.9.139 (6dbba07d9) cannot load the library.tsv file from EasyPQP fcyu  2021-06-02
 

The library.tsv file from EasyPQP is in OpenSWATH format. But when loading it in Skyline, it says "the file F:\dev\msfragger\test_skyline\library.tsv is not a supported spectral library file format.". According to https://skyline.ms/wiki/home/software/BiblioSpec/page.view?name=BlibBuild, it should support OpenSWATH tsv file.

I put an example of library.tsv in the attachment. Could you please take a look when you have time?

Thanks,

Fengchao

 
 
Kaipo responded:  2021-06-02

Hi Fengchao,

This doesn't look like the OpenSWATH files we have seen in the past. Normally we expect these columns:

  • filename
  • RT
  • FullPeptideName
  • Charge
  • m/z
  • decoy
  • aggr_Peak_Area
  • aggr_Fragment_Annotation
  • m_score

and optionally these ones:

  • ProteinName
  • leftWidth
  • rightWidth

Additionally, to build spectral libraries, Skyline requires results files that have some type of probability scores (e.g. q-values) and a way to map results to a spectrum file (e.g. filename and scan number). If you have that information, it should be possible to generate a generic ssl file (https://skyline.ms/wiki/home/software/BiblioSpec/page.view?name=BiblioSpec input and output file formats) that can be used to build a library.

Thanks,
Kaipo

 
fcyu responded:  2021-07-15

Hi Kaipo,

Thanks for your reply.

I am trying another approach to import this library file through File -> Import -> Assay Library. Unfortunately, I get another error says: Precursor m/z 300.8289 does not match the closest possible value 295.4972 (delta = 5.3317), peptide MALLHSGR. Check the Modification tab in the Peptide Settings, the m/z types on the Prediction tab, or the m/z match tolerance on the Instrument tab of the Transition Settings.

The corresponding row is

PrecursorMzProductMzAnnotationProteinIdGeneNamePeptideSequenceModifiedPeptideSequencePrecursorChargeLibraryIntensityNormalizedRetentionTimePrecursorIonMobilityFragmentTypeFragmentChargeFragmentSeriesNumberFragmentLossType
300.828851175.118953y1^1P00505GOT2MALLHSGRM(UniMod:35)ALLHSGR31503.963184-28.72662465768476y11

It seems that Skyline cannot recognize the Unimod notation (UniMod:35).

Could you please tell me how to solve this issue?

Thanks,

Fengchao

 
Kaipo responded:  2021-07-15

Hi Fengchao,

First you will need to ensure that the necessary modifications are added (through Settings > Peptide Settings > Modifications, as the error mentions).

Then, normally you would be able to use "[UniMod:##]" format but it looks like there is a bug preventing this from working. As a workaround for the time being you can use a text editor to find and replace the UniMod strings with the modification masses, e.g.:

  • oxidation: replace "(UniMod:35)" with "[+16.0]"
  • carbamidomethylation: replace "(UniMod:4)" with "[+57.0]"

And retry the import after the replacements have been made.

Thanks,
Kaipo

 
fcyu responded:  2021-07-15

Hi Kaipo,

Thanks for your prompt reply.

Your suggestions work for those modifications now. But I have new errors:
Precursor m/z 351.1602 does not match the closest possible value 350.4883 (delta = 0.6719), peptide EREMAEMR. Check the Modification tab in the Peptide Settings, the m/z types on the Prediction tab, or the m/z match tolerance on the Instrument tab of the Transition Settings.

The row is like

PrecursorMzProductMzAnnotationProteinIdGeneNamePeptideSequenceModifiedPeptideSequencePrecursorChargeLibraryIntensityNormalizedRetentionTimePrecursorIonMobilityFragmentTypeFragmentChargeFragmentSeriesNumberFragmentLossType
351.16016099999996306.15943799999997y2^1P02545LMNAEREMAEMREREMAEMR310000.03.363432645878369y12

The m/z difference (351.1602 vs 350.4883 ) is from isotope errors (we have isotope error 0/1/2 during the search). Could you please provide any suggestions or solutions to this issue?

Another question I would like to ask is that do you have any suggestion regarding the library format that can be natively supported by Skyline? The file I am using is from EasyPQP without any file name or scan name information because those are consensus spectra.

Thanks,

Fengchao

 
fcyu responded:  2021-07-15

Hi Kaipo,

I figured out the isotope error issue. But I have another one (sorry that I have so many questions):
Failed to explain all transitions for SVVLMSHLGRPDGVPMPDK m/z 513.5158 with a single set of modifications

The row is

PrecursorMzProductMzAnnotationProteinIdGeneNamePeptideSequenceModifiedPeptideSequencePrecursorChargeLibraryIntensityNormalizedRetentionTimePrecursorIonMobilityFragmentTypeFragmentChargeFragmentSeriesNumberFragmentLossType
513.515771724.887564b14^2P00558PGK1SVVLMSHLGRPDGVPMPDKSVVLMSHLGRPDGVPM[+15.994900]PDK4110.35082429.329321355011768b214

The b14^2 ion do have 724.8876 Th. I am not sure what causes this error.

Thank in advance for your help.

Best,

Fengchao

 
Kaipo responded:  2021-07-16

Hi Fengchao,

I don't see that line in the tsv - is it from the same one you uploaded earlier?

Thanks,
Kaipo

 
fcyu responded:  2021-07-16

Hi Kaipi,

I am sorry for the confusion. Yes, it is from another library. I put it alone with Skyline files in the attachment.

Thanks again for your time and help,

Fengchao

 
Kaipo responded:  2021-07-28

Hi Fengchao,
Skyline is having trouble with certain sequences that have two modifications on the same amino acid, e.g.:
[+42.010600]C[+57.021464]DAFVGTWK
which Skyline interprets as
C[+42.010600][+57.021464]DAFVGTWK
There are 35 lines like this in the uploaded file - you can replace these instances with a single modification that combines the formula of both modifications (something like C[+99.032064]DAFVGTWK and create a corresponding modification in Skyline).

There does also appear to be some sequences missing the carbamidomethyl modification, which will result in errors like this:
Precursor m/z 372.4297 does not match the closest possible value 358.1743 (delta = 14.2554), peptide CTAGTLHNLSHHR. Check the Modification tab in the Peptide Settings, the m/z types on the Prediction tab, or the m/z match tolerance on the Instrument tab of the Transition Settings.
After adding a +57 mass shift to the cysteines the errors went away. Hope that helps.

Thanks,
Kaipo