Creation of a Skyline spectral library from Kojak XL-MS search results #1
Comments
|
Hi Alessio, thanks for your message. I got your email as well, apologies for taking so long to respond. I am happy to help however I can. Just so I understand: You are able to create a Proxl XML file from Kojak results, both with and without using Percolator. When you try to create a spectral library in Skyline, the Proxl XML file you made from Kojak-only results works, but trying to do so with the Proxl XML created from Kojak+Percolator gives an error. Are any error messages produced by Skyline that you can paste in here? |
|
search_proxl.xml
<https://drive.google.com/file/d/1X51R0eyxTG2AN6EfAWEYZz9tCqPqKEl9/view?usp=drive_web>
Hi Michael, no worries at all. So to clarify this (it was a very long email
I apologize), I am able to run Kojak XL-MS searches and doing validation
with Percolator (from what I see in the xml Kojak output prior to
Percolator, there's already an e-value probably meaning there's already a
statistic control on CSMs level). I then tried to use proXL converter for
Kojak for 1) Kojak-only results and 2) Kojak results with percolator
validated files (both .txt and .xml formats-->the latter is the one
provided in the exemplary help function when you run the jar file in a
windows cmd terminal).
For 1) I am able to convert the results into an XML file which seems
reasonable to me (I attached it as a Google Drive link so that you can have
a look at it yourself) and the script ran successfully with no error saying
that conversion was properly done.
[image: image.png]
Unfortunately, when I try to create a spec lib in Skyline (Via
Settings-->Peptide Settings-->Library) and then trying to build a library
using the .xml files in the same folder of my mzml files, the ProXL
converted XML output file does not appear and if I force Skyline to open it
it give me this error:
[image: image.png]
For 2) either trying with .txt or .xml files didn't work and gives me this
error:
*lineNumber: 1; columnNumber: 1; Content is not allowed in prolog*.;
probably because the first line of the percolator .xml file does not start
with the usual <?xml version="1.0" encoding="UTF-8"> (??). Even trying to
put that in manually didn't solve the problem and I still get the same
error. Also changing UTF didn't change anything.
I am currently stuck at this point, so if you could try to help me out with
that it would be awesome.
PS: If you need me to send you the configuration file or the other files
please let me know.
Best regards
Alessio
Il giorno gio 14 dic 2023 alle ore 19:28 Michael Riffle <
***@***.***> ha scritto:
… Hi Alessio, thanks for your message. I got your email as well, apologies
for taking so long to respond. I am happy to help however I can.
Just so I understand: You are able to create a Proxl XML file from Kojak
results, both with and without using Percolator. When you try to create a
spectral library in Skyline, the Proxl XML file you made from Kojak-only
results works, but trying to do so with the Proxl XML created from
Kojak+Percolator gives an error.
Are any error messages produced by Skyline that you can paste in here?
—
Reply to this email directly, view it on GitHub
<#1 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/A6LHTWQU2HNVOH7LYRQK5L3YJNAOHAVCNFSM6AAAAABAPJ7MBWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTQNJWGM3TENBYGE>
.
You are receiving this because you authored the thread.Message ID:
***@***.***>
|
|
|
1. From my email I can see all the attachments
2. java -jar kojak2ProxlXML.jar -o search.proxl.xml -l dsso -k
my_kojak_output_file -c Kojak.conf -N (Kojak-only results)
java -jar kojak2ProxlXML.jar -o search.proxl.xml -l dsso -k
my_kojak_output_file.xml -c Kojak.conf -x perco-inter.xml -x
perco-intra.xml -x perco-loop.xml -x perco-single.xml (Kojak-Percolator).
The java executable was in the same folder as all the files needed for
doing the conversion.
For the linker, I put the keyword # linker:dsso below the cross-linker
modifications in the kojak parameter file so that it would be recognized.
Il giorno gio 14 dic 2023 alle ore 20:31 Michael Riffle <
***@***.***> ha scritto:
…
1.
The image of the error from Skyline didn't come through.
2.
Can you provide the exact command line command you are using to run
both percolator and the Proxl XML converter?
—
Reply to this email directly, view it on GitHub
<#1 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/A6LHTWUBPLHNAUGEBY7RK73YJNHXJAVCNFSM6AAAAABAPJ7MBWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTQNJWGQ2TOMBVGI>
.
You are receiving this because you authored the thread.Message ID:
***@***.***>
|
|
The error in Skyline is : the output.xml file is not a valid library input
file
Il giorno gio 14 dic 2023 alle ore 21:08 Alessio Di Ianni <
***@***.***> ha scritto:
… 1. From my email I can see all the attachments
2. java -jar kojak2ProxlXML.jar -o search.proxl.xml -l dsso -k
my_kojak_output_file -c Kojak.conf -N (Kojak-only results)
java -jar kojak2ProxlXML.jar -o search.proxl.xml -l dsso -k
my_kojak_output_file.xml -c Kojak.conf -x perco-inter.xml -x
perco-intra.xml -x perco-loop.xml -x perco-single.xml (Kojak-Percolator).
The java executable was in the same folder as all the files needed for
doing the conversion.
For the linker, I put the keyword # linker:dsso below the cross-linker
modifications in the kojak parameter file so that it would be recognized.
Il giorno gio 14 dic 2023 alle ore 20:31 Michael Riffle <
***@***.***> ha scritto:
>
> 1.
>
> The image of the error from Skyline didn't come through.
> 2.
>
> Can you provide the exact command line command you are using to run
> both percolator and the Proxl XML converter?
>
> —
> Reply to this email directly, view it on GitHub
> <#1 (comment)>,
> or unsubscribe
> <https://github.com/notifications/unsubscribe-auth/A6LHTWUBPLHNAUGEBY7RK73YJNHXJAVCNFSM6AAAAABAPJ7MBWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTQNJWGQ2TOMBVGI>
> .
> You are receiving this because you authored the thread.Message ID:
> ***@***.***>
>
|
What are perco-inter.xml, perco-intra.xml etc. Were those produced by percolator? What is the command you used to run percolator? Can you attach one of those perco-*.xml files? I just want to confirm they're in the expected format. |
|
Yes, these were produced by Percolator. I've used the following command:
*C:\MyFolder\Percolator.exe DataFile1.pin > DataFile1-validated.txt (or
.xml).*
The pin (percolator input files) were generated by putting the parameter
export_percolator = 1 in the kojak.conf file. Attached you can find a pin
file and a produced percolator file after validation.
Il giorno gio 14 dic 2023 alle ore 21:24 Michael Riffle <
***@***.***> ha scritto:
… erco-inter.x
What are perco-inter.xml, perco-intra.xml etc. Were those produced by
percolator? What is the command you used to run percolator? Can you attach
one of those perco-*.xml files? I just want to confirm they're in the
expected format.
—
Reply to this email directly, view it on GitHub
<#1 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/A6LHTWQCCVDPYXVLUQTPVZLYJNN6DAVCNFSM6AAAAABAPJ7MBWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTQNJWGUZDSOJXHE>
.
You are receiving this because you authored the thread.Message ID:
***@***.***>
|
|
Aha, I see the issue (at least one of the issues): You have to give Percolator another flag to output XML, changing the name of the file from .txt to .xml won't work. Here is how to run it:
This will create two files, the XML file and the normal text file. The converter needs to have the XML file. |
|
Ok, I was doing doing this command as suggested on the Kojak website
C:\Percolator\Percolator.exe
DataFile1-percolator.txt > DataFile1-validated.txt (there was no indication
on how to do it with the .pin file). I will do it this way from now on. But
this will not solve the problem of the speclib creation in Skyline without
using Kojak-only results (no perc files needed but only the kojak output
file and the conf file). I've tried now to do the conversion in case of
Kojak+Percolator files using the new command you said and it still gives an
error. Thanks a lot for your help though, hope you'll find the mistake here.
Il giorno gio 14 dic 2023 alle ore 22:50 Michael Riffle <
***@***.***> ha scritto:
… Aha, I see the issue (at least one of the issues):
You have to give Percolator another flag to output XML, changing the name
of the file from .txt to .xml won't work.
Here is how to run it:
C:\MyFolder\Percolator.exe -X DataFile1-validated.xml DataFile1.pin >
DataFile1-validated.txt
This will create two files, the XML file and the normal text file. The
converter needs to have the XML file.
—
Reply to this email directly, view it on GitHub
<#1 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/A6LHTWV7HT3VCMILFEWPEM3YJNYBJAVCNFSM6AAAAABAPJ7MBWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTQNJWG4YDMMBTGI>
.
You are receiving this because you authored the thread.Message ID:
***@***.***>
|
|
We might need to find a better form of communication, zoom maybe.
By default, Percolator doesn't make the XML required for the Proxl XML converter. You have to use the -X flag to generate the XML that it needs. The error you're seeing is because a non-XML document is being given to the Proxl XL converter. Sequence of events should be roughly:
I will make a different post about the Skyline part. |
|
The Skyline issue: For its cross-linking support, Skyline requires that the data be represented as Proxl XML, which you know. Internally it uses a program called Bibliospec, that is part of Proteowizard, to create the spectral library from the ProxlXML file and the spectral files (e.g. mzML or MGF files). Since ProxlXML is a format that can contain data from any program (e.g. Kojak or MeroX), and since Bibliospec can't know about every possible program that does cross-linking, it looks for q-values to be present in the data (or some form of probability). The developers of Proteowizard don't want to support native scores that come out of programs, but want to stick to probability-based scores, like q-values, when creating spectral libraries. Bibliospec doesn't know how to read Proxl XML files from Kojak-only--it doesn't know about e-values. I don't know if they'd be willing to support that or not. I'm not a member of the Proteowizard or Skyline teams. Skyline has a support forum here: https://skyline.ms/project/home/support/begin.view? I know they have been very responsive in the past about this kind of issue. It's possible that Skyline will support the Kojak+Percolator data, since it contains q-values. But I don't know this for sure. |
|
Thanks for the detailed explanation. It might be good to have a meeting (Zoom or whatever platform you’re comfortable with).On Dec 15, 2023, at 20:14, Michael Riffle ***@***.***> wrote:
The Skyline issue:
For its cross-linking support, Skyline requires that the data be represented as Proxl XML, which you know. Internally it uses a program called Bibliospec, that is part of Proteowizard, to create the spectral library from the ProxlXML file and the spectral files (e.g. mzML or MGF files). Since ProxlXML is a format that can contain data from any program (e.g. Kojak or MeroX), and since Bibliospec can't know about every possible program that does cross-linking, it looks for q-values to be present in the data (or some form of probability). The developers of Proteowizard don't want to support native scores that come out of programs, but want to stick to probability-based scores, like q-values, when creating spectral libraries.
Bibliospec doesn't know how to read Proxl XML files from Kojak-only--it doesn't know about e-values. I don't know if they'd be willing to support that or not. I'm not a member of the Proteowizard or Skyline teams. Skyline has a support forum here: https://skyline.ms/project/home/support/begin.view? I know they have been very responsive in the past about this kind of issue.
It's possible that Skyline will support the Kojak+Percolator data, since it contains q-values. But I don't know this for sure.
—Reply to this email directly, view it on GitHub, or unsubscribe.You are receiving this because you authored the thread.Message ID: ***@***.***>
|
|
Hi Michael, I tried your suggestion for the Kojak+Percolator workflow with
the new command you suggested but when it comes to point 3 (ProXL
conversion), now I get the error:
java.lang.Exception: Cannot parse scanNumber '211026EWas01_E1', psmIdString
= 'T-211026EWas01_E1-26916-82.26'. Would it be ok for you to have a meeting
for this? For the kojak only results I will try to directly reach out the
Skyline support for help regarding that issue.
Best
Alessio
Il giorno ven 15 dic 2023 alle ore 22:01 Alessio Di Ianni <
***@***.***> ha scritto:
… Thanks for the detailed explanation. It might be good to have a meeting
(Zoom or whatever platform you’re comfortable with).
On Dec 15, 2023, at 20:14, Michael Riffle ***@***.***>
wrote:
The Skyline issue:
For its cross-linking support, Skyline requires that the data be
represented as Proxl XML, which you know. Internally it uses a program
called Bibliospec, that is part of Proteowizard, to create the spectral
library from the ProxlXML file and the spectral files (e.g. mzML or MGF
files). Since ProxlXML is a format that can contain data from any program
(e.g. Kojak or MeroX), and since Bibliospec can't know about every possible
program that does cross-linking, it looks for q-values to be present in the
data (or some form of probability). The developers of Proteowizard don't
want to support native scores that come out of programs, but want to stick
to probability-based scores, like q-values, when creating spectral
libraries.
Bibliospec doesn't know how to read Proxl XML files from Kojak-only--it
doesn't know about e-values. I don't know if they'd be willing to support
that or not. I'm not a member of the Proteowizard or Skyline teams. Skyline
has a support forum here:
https://skyline.ms/project/home/support/begin.view? I know they have been
very responsive in the past about this kind of issue.
It's possible that Skyline will support the Kojak+Percolator data, since
it contains q-values. But I don't know this for sure.
—
Reply to this email directly, view it on GitHub
<#1 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/A6LHTWVTR7HTZDGDQRN45LLYJSOTBAVCNFSM6AAAAABAPJ7MBWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTQNJYGM3TAMZYHA>
.
You are receiving this because you authored the thread.Message ID:
***@***.***>
|
|
Hi Michael,
here Alessio again. I hope you had a good start into the new year. As said
previously, I tried your suggestions for the Kojak+Percolator workflow but
it gave me issues when trying to convert the Percolator .xml results into
ProXL .xml results. Could I send you my result files so you can try
yourself and see if the ProXL conversion works?? Alternatively, could we
also have a Zoom call to talk about this?
Best
Il giorno mar 19 dic 2023 alle ore 10:27 Alessio Di Ianni <
***@***.***> ha scritto:
… Hi Michael, I tried your suggestion for the Kojak+Percolator workflow with
the new command you suggested but when it comes to point 3 (ProXL
conversion), now I get the error:
java.lang.Exception: Cannot parse scanNumber '211026EWas01_E1',
psmIdString = 'T-211026EWas01_E1-26916-82.26'. Would it be ok for you to
have a meeting for this? For the kojak only results I will try to directly
reach out the Skyline support for help regarding that issue.
Best
Alessio
Il giorno ven 15 dic 2023 alle ore 22:01 Alessio Di Ianni <
***@***.***> ha scritto:
> Thanks for the detailed explanation. It might be good to have a meeting
> (Zoom or whatever platform you’re comfortable with).
>
>
> On Dec 15, 2023, at 20:14, Michael Riffle ***@***.***>
> wrote:
>
>
>
> The Skyline issue:
>
> For its cross-linking support, Skyline requires that the data be
> represented as Proxl XML, which you know. Internally it uses a program
> called Bibliospec, that is part of Proteowizard, to create the spectral
> library from the ProxlXML file and the spectral files (e.g. mzML or MGF
> files). Since ProxlXML is a format that can contain data from any program
> (e.g. Kojak or MeroX), and since Bibliospec can't know about every possible
> program that does cross-linking, it looks for q-values to be present in the
> data (or some form of probability). The developers of Proteowizard don't
> want to support native scores that come out of programs, but want to stick
> to probability-based scores, like q-values, when creating spectral
> libraries.
>
> Bibliospec doesn't know how to read Proxl XML files from Kojak-only--it
> doesn't know about e-values. I don't know if they'd be willing to support
> that or not. I'm not a member of the Proteowizard or Skyline teams. Skyline
> has a support forum here:
> https://skyline.ms/project/home/support/begin.view? I know they have
> been very responsive in the past about this kind of issue.
>
> It's possible that Skyline will support the Kojak+Percolator data, since
> it contains q-values. But I don't know this for sure.
>
> —
> Reply to this email directly, view it on GitHub
> <#1 (comment)>,
> or unsubscribe
> <https://github.com/notifications/unsubscribe-auth/A6LHTWVTR7HTZDGDQRN45LLYJSOTBAVCNFSM6AAAAABAPJ7MBWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTQNJYGM3TAMZYHA>
> .
> You are receiving this because you authored the thread.Message ID:
> ***@***.***>
>
>
|
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Dear proXL team,
I am trying to use Kojak 2.1.0 for XL-MS data analysis. I am currently doing it from the command line in combination with Percolator validation; all the runs look fine and I was able to see the data with the Kojak SpectrumViewer executable. I then tried to convert the kojak.txt output file into a proXL XML file for subsequent spectral library creation and manual validation in Skyline but didn't manage to read the XML file. Shortly, using the command for ProXL conversion with Percolator .xml files gives me error, while the conversion without using percolator validated files is successfull and it seems to generate a good XML output file but for some strange reason I am not able to read it in Skyline. Could you help him out to convert the data in both scenarios so that I can import everything into Skyline afterwards?
The text was updated successfully, but these errors were encountered: