PeptideShaker project generation error and Santa wish!

[Juan Camilo Rojas Echeverri]

Dear support,

I am having some issues trying to load my SearchGUI (3.3.17) results into a PeptideShaker project (1.16.42). I am attaching the text log.

My data is from a Waters Synapt G2-Si instrument acquired in DDA mode. I have to rely on either PLGS or Progenesis to produce well formatted .mgf files.

I was able to successfully create a project with the export from Progenesis. However, when I try using the same data exported from PLGS to .mzML (which for some reason lacks the MS1 data...) and then MSConverted to .mgf  (I know it seems redundant, but I find the direct .mgf export TITLE notation useless) export I get this error:

"An error occured: Not enough representative ptm sites found"

If it wasn't because I have already been able to confirm many PTMs with the Progenesis .mgf generated project and parallely with PEAKs I would have thought it was a problem with the quality of my data, but I doubt that is the case.

I had this problem before with some data that had very few MSMS, but after disabling the PTM scoring and making wider import filters at least I could load my peptide IDs so I could confirm manually. Nevertheless, this time that approach didn't work.

My search took way too many days and would like to know if somehow this could be salvaged into a PeptideShaker project before attempting another long search with it...

Santa wish: could you please please device a way of working with Waters .RAW files? :'(

I don't want to rely on a Progenesis black box that preprocesses my data in ways I can't know since they are very inflexible when it comes to explaining their decision making on the preprocessing and more importantly there are VERY limited options for user defined parameters.

I have also tried with MSConvert both with the GUI and the command line application to try to create well formatted .mgf, but either I get a file that has no charge determination or charge of "2+, 3+, 4+, 5+". Theoretically, for TOF data it should also be possible use a filter that performs precursor recalculation (which is also very important) using the MS1 data and charge determination, but this crashes the app saying the MS detector is not compatible or something like that.

At some poing I'll try to contact ProteoWizard support to address those last two, but would be interested to hear your suggestions.

Sorry for the long error report and christmas wish list.

Sincerely,

Juan C. Rojas

[Harald Barsnes]

Hi Juan,

This sounds like an issue that is hopefully solved in the new beta releases (https://groups.google.com/forum/#!topic/peptide-shaker/1ecY0IyMOBM).

It would therefore be great if you could try loading the data in the new version of PeptideShaker to see if that takes care of the error? You do not have redo the search, but you will have to recreate the search parameter file in PeptideShaker before loading the data.

If you are using more than one search engine, you may also try loading the results of only one of the search engines in (the either the old or the new version of) PeptideShaker at the time. This could potentially avoid the issue as well, as some of the search engines are more prone to creating PTM issues than others. I recommend starting with only X! Tandem.

As for the Santa wish regarding conversion of Waters raw files, I'm afraid we do not have the resources to look into this ourselves. It's also a bit outside our area of expertise. Your idea if contacting ProteoWizard therefore seems like the best available option. Unless someone else on this list have other suggestions?

Best regards,

Harald

[Marc Vaudel]

Hi Juan,

I have discussed your Santa wish with some of the best of his bioinformatics crew and I am afraid they did not find an easy answer. The easiest would be to get ProteoWizzard to support your files. Maybe your instrument is not implemented yet or there is a problem with the code, in any case it would be good to contact them. We are working on better support of raw and mzML data, but I am afraid it is not realistic to expect a release before 2021.

Apologies that we cannot help more at this stage,

Marc

--
You received this message because you are subscribed to the Google Groups "PeptideShaker" group.
To unsubscribe from this group and stop receiving emails from it, send an email to peptide-shake...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/peptide-shaker/8bec3e6f-390c-4090-b37b-eaa072f24a96%40googlegroups.com.

[Juan Camilo Rojas Echeverri]

Harald and Marc,

Thank you very much for your responses. It's a shame that I was not good enough to be in Santa's good list last year :/

I performed my searches with: X!Tandem, MS-GF+ and MSAmanda with SearchGUI 3.3.17. I tried Harald's suggestions and after trying several combinations of loading the data in PeptideShaker-1.16.42 and PeptideShaker-2.0.0-beta this is what I found.

- PeptideShaker-1.16.42

    - X!Tandem results: No error

    - MS-GF+ results: No error

    - MS Amanda results: "An error occurred: Not enough representative ptm sites found."

    - X!Tandem and MS-GF+ results: No error

    - X!Tandem, MS-GF+, and MS Amanda results: "An error occurred: Not enough representative ptm sites found."

- PeptideShaker-2.0.0-beta

    - MS Amanda results: No error

    - X!Tandem, MS-GF+, and MS Amanda results (trial #1): No error was given, but the software froze once it reached the "Computing PSM probabilities" and didn't complete. There was a log of some Java error (memory issues?)

    - X!Tandem, MS-GF+, and MS Amanda results (trial #2): No error was given. SUCCESS! :)

It seems to me the results given by MS Amanda were the problem in PeptideShaker-1.16.42 and somehow the beta version is being able to manage the issue. I'm guessing you have done a lot of new things with the beta version, but could you guess what specifically allowed the beta version manage the error?

With regards to the data converter, thanks for the head-ups. I'll go bug ProteoWizard people now.

Thank you for your time and quick responses.

Sincerely,

JC

To view this discussion on the web visit https://groups.google.com/d/msgid/peptide-shaker/CAE1e1duKWFFtd6KdcXGvrD3k%2BMpCwRz1b-ORCCLXHXKgkjBhsA%40mail.gmail.com.

[Harald Barsnes]

Hi Juan,

> I'm guessing you have done a lot of new things with the beta version, but could you guess what specifically allowed the beta version manage the error?

We more or less changed all of the back end code in the new beta versions, so it's difficult to pinpoint exactly which change ensured that you are able to load the data. Probably it's a combination of different improvements. :)

Best regards,

Harald