ERROR: No spectra were found for the new library - since updating to v 21.2

support
ERROR: No spectra were found for the new library - since updating to v 21.2 Francis Beaudry  2022-04-28
 
Hi,
 
Since I updated to 21.2, I have a very difficult time with Skyline. Sometimes it works, but most of the time it doesn’t work. Currently, we are trying to process DDA datasets and use the option DDA Raw to search and build the library. Attached, you will find the message we are getting 9 times out of 10 trials using the same raw files and search setup. Weird … Can you help ??
 
Regards,
Francis
 
---------------------------
Skyline
---------------------------
ERROR: No spectra were found for the new library.
 
Command-line: C:\Program Files\Skyline\BlibBuild -s -A -H -o -c 0.95 -i 07ADT_27avr_std_3 -S "C:\Users\francisb\AppData\Local\Temp\tmp8B1F.tmp" "C:\Users\francisb\Documents\FB-PROJECTS\ADT\07ADT_Validation_tests\27avr22\Skyline\07ADT 27avr std 3.redundant.blib"
Working directory: C:\Users\francisb\Documents\FB-PROJECTS\ADT\07ADT_Validation_tests\27avr22
---------------------------
OK More Info
---------------------------
System.IO.IOException: ERROR: No spectra were found for the new library.
 
Command-line: C:\Program Files\Skyline\BlibBuild -s -A -H -o -c 0.95 -i 07ADT_27avr_std_3 -S "C:\Users\francisb\AppData\Local\Temp\tmp8B1F.tmp" "C:\Users\francisb\Documents\FB-PROJECTS\ADT\07ADT_Validation_tests\27avr22\Skyline\07ADT 27avr std 3.redundant.blib"
Working directory: C:\Users\francisb\Documents\FB-PROJECTS\ADT\07ADT_Validation_tests\27avr22
   à pwiz.Common.SystemUtil.ProcessRunner.Run(ProcessStartInfo psi, String stdin, IProgressMonitor progress, IProgressStatus& status, TextWriter writer, ProcessPriorityClass priorityClass) dans C:\proj\skyline_21_2_x64\pwiz_tools\Shared\Common\SystemUtil\ProcessRunner.cs:ligne 149
   à pwiz.BiblioSpec.BlibBuild.BuildLibrary(LibraryBuildAction libraryBuildAction, IProgressMonitor progressMonitor, IProgressStatus& status, String& commandArgs, String& messageLog, String[]& ambiguous) dans C:\proj\skyline_21_2_x64\pwiz_tools\Shared\BiblioSpec\BlibBuild.cs:ligne 201
   à pwiz.Skyline.Model.Lib.BiblioSpecLiteBuilder.BuildLibrary(IProgressMonitor progress) dans C:\proj\skyline_21_2_x64\pwiz_tools\Skyline\Model\Lib\BiblioSpecLiteBuilder.cs:ligne 157
---------------------------
 
 
matt.chambers42 responded:  2022-04-28
Hi Francis,

Huh, you mean you can repeat the same search that fails most of the time but eventually it will work, without changing any settings? Can you tell us what search settings you are using and what kind of raw input format?

Also please try the latest Skyline-daily to see if there's an improvement there. If it's not fixed in daily, if you could send us one of your problematic files (and FASTA) so we can reproduce locally, that'll help us find the problem quickly. Thanks for the report and your patience!
 
Francis Beaudry responded:  2022-04-28
Dear Matt

Thank you for your quick response, it is much appreciated. I have installed and reinstalled Skyline. I also installed the admin version. I haven't tried yet the Skyline-daily. I will try this weekend.

Attached (zip file), you will find the raw file, the FASTA files and two skyline trials (1-successful, 2-unsuccessful).

The RAW data is from a DDA top 10 analysis performed on a QE +. The samples contained only two standard proteins (Bovin CYC and equine MYB) for testing.

Regards,
Francis
 
matt.chambers42 responded:  2022-04-28
Hi Francis, perhaps the zip fil was too big to attach? If so you can upload it to https://skyline.ms/project/home/support/file%20sharing
 
Francis Beaudry responded:  2022-04-28
Dear Matt

Thank you for your quick response, it is much appreciated. I have installed and reinstalled Skyline. I also installed the admin version. I haven't tried yet the Skyline-daily. I will try this weekend.

Attached (zip files), you will find the raw file, the FASTA files and two skyline trials (1-successful, 2-unsuccessful).

The RAW data is from a DDA top 10 analysis performed on a QE +. The samples contained only two standard proteins (Bovin CYC and equine MYB) for testing.

Regards,
Francis
 
Nick Shulman responded:  2022-04-28
Francis,

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

(The URL that Matt posted was almost correct but was missing "/begin.view")
-- Nick
 
Francis Beaudry responded:  2022-04-28
Dear Matt

Thank you for your quick response, it is much appreciated. I have installed and reinstalled Skyline. I also installed the admin version. I haven't tried yet the Skyline-daily. I will try this weekend.

Attached (zip files), you will find the FASTA files and two skyline trials (1-successful, 2-unsuccessful). I am unable to send the Thermo RAW file. The zip file is > 50MB limit. I am not sure if these file will be useful without the raw file.

The RAW data is from a DDA top 10 analysis performed on a QE +. The samples contained only two standard proteins (Bovin CYC and equine MYB) for testing.

Regards,
Francis
 
Francis Beaudry responded:  2022-04-28
Dear Matt,

I just used the link to download the zip file (skyline.zip; description: Skyline-Error-FB).

Regards,
Francis
 
matt.chambers42 responded:  2022-05-02
I was not able to reproduce any sporadic failure. With given search settings, I either got some results or I didn't. Although it did give me an idea for a feature to remember/re-use the search settings from the previous import job (even if it failed and is not recorded in the audit log)!

Is this 2 standard protein kind of run the only one where you are getting failures? It can be tricky for search engines to get decent good FDR/Q-value scores when searching a huge database against a small set of potential true positives. I think that's what's happening here. If you set the cutoff score to .5 or so, then you may see some hits. You can also try one of the other search engines, but they all have this issue to some degree. It's a direct consequence of correcting scores for multiple comparisons. But there may be some other factors in play because I get somewhat better results when I widened the search space a bit (higher precursor and fragment tolerance, e.g. 25 ppm for both). Has there been some kind of labeling or modification done to the proteins? Because 5 to 7 peptides does seem pretty low for such a simple sample.
 
Francis Beaudry responded:  2022-05-02
Dear Matt,
Thank you for the follow-up. We use these proteins for QC and optimization purposes. In these samples, there is no labeling. In PD 2.4, we are able to ID 28 peptides with high confidence.

Until the update to Skyline 21.2, I had no issues with these samples or other similar tests we run every week. We use the option of separate FASTA for targeted analysis. This option would, I presume, alleviate some of the tricky situation. I also experience similar problems with untargeted DDA experiments on cell extracts or other biological samples. Normally, we can ID/Quan 1500 to 3000 proteins from these samples.

As you suggest, setting a cutoff score to .5 seem to solve the problem. I just tried it twice and it works.
 
matt.chambers42 responded:  2022-05-02
28 peptides? That's pretty far from the 7 I'm getting (I tested both MSAmanda and MSFragger). Are you using the same equus FASTA? Also are you somehow searching both FASTAs with PD but only one FASTA with Skyline?