OpenMS/OpenMS

The codebase of the OpenMS project

Julianus Pfeuffer

@jpfeuffer

@kairenchen721 We are testing for a range of common decoy strings. Yes apparently none of them is present. The accession is only used if --picked_FDR is used/true. Otherwise it uses annotations from PeptideIndexer (i.e. the target_decoy meta values)

1 reply

But you also seem to have another problem. Apparently, you have a lot of unconnected nodes in your graph.

This could be due to the prefiltering or due to wrong creation of the input file.

Douglas McCloskey

@dmccloskey

Does anyone know if the AccurateMassSearchEngine modifies the db_mapping_file or db_struct_file identifiers? Specifically, IDs of the form "glc__D_c" appear to be converted to "glc".

1 reply

Douglas McCloskey

@dmccloskey

iJO1366_mapping_mod.tsv

iJO1366_struct_mod.tsv

@matteopilz I have attached the database files I was working with. Please let me know if I can provide anything else.

1 reply

Douglas McCloskey

@dmccloskey

01_BOP_1000_sucB_ALE #1_SS-NT-P-E-M_1_LowNeg10x-2_4566_SS-NT-P-E-M_1900-01-01_000000.featureXML

@matteopilz I have attached a single example. Please let me know if any additional examples can be useful.

5 replies

Roger Olivella

@rolivella

Hi, I'm working with the CometAdapter and I have some issues. When I search a light file with some variable modifications it works fine. But when I use a quite big file (for instance a 2.2GB mzML) and more that three variable modifications the search takes ages and does not finish. I tried the same search without using the CometAdapter, I mean, by running the comet.linux.exe and the search ends in 1h aprox. Do you have any idea on why the CometAdapter is taking so long? I can upload all the files in a Dropbox and the command line instructions I'm running if you want to debug it. Thanks!

Samuel Wein

@poshul

Hi @rolivella, can you open a bug report over at https://github.com/OpenMS/OpenMS/issues ? It's definitely something that is worth our team following up on

Julianus Pfeuffer

@jpfeuffer

And the first thing to do would probably be to check the param file written by OpenMS (unless you used exactly that one for your search with the executable). And I assume you also used exactly the same executable?

Roger Olivella

@rolivella

Thanks @poshul and @jpfeuffer I'll open an issue. Yes, I'm using the same exe.

Kai Ren Chen

@kairenchen721

Hi @jpfeuffer, I would like to ask some questions about the expected results of epifany. So both figure (please see pdf below) shows a component, the top row are peptides with their qvalue and pep, the bottom row is are the protein with their qvalue. The figure with the higher qvalue for protein is the one I ran with more decoy

I was wondering if it is supposed to happen that when I added more decoy, the qvalue increases.

I was also wondering that if all protein are supposed to be similar in qvalue?

Thank you

Julianus Pfeuffer

@jpfeuffer

If you add significantly more decoys than targets, you need to correct the resulting q-values for this bias.

69 replies

Oops I deleted instead of edited my last comment.

I need to see the decoys to say if it looks correct. The results are definitely plausible. Q-values only change when a decoy appears in the Posterior Probability ranking. So before the first decoy, every target has a q-value of around 0 and between two decoys, all proteins will have the same q-value. Q-value is only for filtering, not really for a distinctive score.

ellenweii

@ellenweii

Hello, I am trying to use pyopenms to search for PTMs in a proteomics dataset. I'm having issues trying to find info related to this in the documentation https://pyopenms.readthedocs.io/en/latest/index.html . So far it looks like the ModifiedPeptideGenerator class might be what I'm looking for, but I am unable to access the functions in the documentation http://www.openms.de/current_doxygen/html/classOpenMS_1_1ModifiedPeptideGenerator.html . Any help would be greatly appreciated, thank you!

Timo Sachsenberg

@timosachsenberg

Hi @ellenweii can you provide me some details on what you want to achieve? This would make it easier for me to give better advice. Feel free to write to me in private/direct message.

I only used the ModifiedPeptideGenerator in C++. There you first retrieve the list of modifications given a list (e.g., ["Oxidiation (M)", "Phospho (S)"] and the method getModifications()

https://github.com/OpenMS/OpenMS/blob/develop/src/pyOpenMS/pxds/ModifiedPeptideGenerator.pxd#L24

and then you apply the modifications using applyVariableModifications

which fills the all_modified_peptides list

once you have the modified amino acid sequences you can use e.g., the TheoreticalSpectrumGenerator to generate in silico spectra

Timo Sachsenberg

@timosachsenberg

I might have found the issue

trying to come up with a fix

Shubham Gupta

@shubham1637

Hi, I am trying to run OpenSWATH on timsTOF DIA data. I converted DIA files with the latest MSConvert. I am using OpenMS docker image ghcr.io/openms/openms-executables:2.8.0. When I ran OpenSWATH the first time, there was an error error message: Requested ion mobility extraction but no ion mobility array found. . The problem seems to be due to header of IM array.
Currently, MSConvert writes array header as

<cvParam cvRef="MS" accession="MS:1003006" name="mean inverse reduced ion mobility array" value="" unitCvRef="MS" unitAccession="MS:1002814" unitName="volt-second per square centimeter"/>

I changed the text to

<cvParam cvRef="MS" accession="MS:1003006" name="non-standard data array" value="Ion Mobility" unitCvRef="MS" unitAccession="MS:1002814" unitName="volt-second per square centimeter"/>

so that value is Ion Mobility as OpenMS parser looks for this keyword
https://github.com/OpenMS/OpenMS/blob/c2555c3278de34a5f261df9a0915f42a8bfada3c/src/openswathalgo/include/OpenMS/OPENSWATHALGO/DATAACCESS/DataStructures.h#L269

However, I am still getting the same error.

---------------------------------------------------
FATAL: uncaught exception!
---------------------------------------------------
last entry in the exception handler: 
exception of type IllegalArgument occurred in line 332, function void OpenMS::ChromatogramExtractorAlgorithm::extractChromatograms(OpenSwath::SpectrumAccessPtr, std::vector<boost::shared_ptr<OpenSwath::OSChromatogram> >&, const std::vector<OpenMS::ChromatogramExtractorAlgorithm::ExtractionCoordinates>&, double, bool, double, const OpenMS::String&) of /OpenMS/src/openms/source/ANALYSIS/OPENSWATH/ChromatogramExtractorAlgorithm.cpp
error message: Requested ion mobility extraction but no ion mobility array found.
---------------------------------------------------

Does anyone what could be going wrong?

KyowonJeong

@KyowonJeong

Is this the same problem as in Roestlab/dia-pasef#31 ?

Shubham Gupta

@shubham1637

Thanks Jeong, It is similar issue. If I use convertTDFtoMzML, it doesn't include ion-mobility array in exported mzML. MSConvert outputs ion mobility array but with a different header "mean inverse reduced ion mobility array".
However, Josh seems to have it fixed in this PR: OpenMS/OpenMS#6234

KyowonJeong

@KyowonJeong

Great! Hope this can resolve the issue

Kai Ren Chen

@kairenchen721

@jpfeuffer

Kai Ren Chen

@kairenchen721

oh, it seems like I enter the wrong pep values for input decoy component

(For the figure only

(for the figure only)

Kai Ren Chen

@kairenchen721

@jpfeuffer

single_component.idXML
before epifany

component_single_output.idXML
after epifany

Andrew Culhane

@drewculhane

Hi All! does anyone know if there any key parameters (potentially inside, but not limited to scope of FeatureFinderMetabo) that affect expectations for isotopologue distributions? Currently trying to find isotopologue series within my .mzXML which are NOT natural abundance for C12:C13, and are instead based on 95:5 labeling. Wondering if there is some parameter that is slipping past my radar which is defaulting to search only for natural abundance, and thats why im not identifying my labeled series.

Julianus Pfeuffer

@jpfeuffer

Hi, the parameter is called "isotope_filtering_model" and should be set to "none" in your case. There is no built-in support for own isotopic distributions yet, so you have to disable filtering.

Andrew Culhane

@drewculhane

Thanks @jpfeuffer , we have actually already set "isotope_filtering_model" to "none", I was curious if there might be some less obvious parameters that are impeding mass series matching that people might know of

Timo Sachsenberg

@timosachsenberg

Did you also set the mz_scoring_13C parameter to true? this might help in assembly of features

Andrew Culhane

@drewculhane

@timosachsenberg Thanks! We currently have it set to false. We're using a mass target list.. may play around to see what adjusting this parameter does...

Julianus Pfeuffer

@jpfeuffer

To be honest, I am not sure if that parameter helps. It basically says that only C13 mass differences are allowed between isotopic traces.

I would rather investigate a specific feature that you are not finding and see why it might not be picked up

rishi2802

@rishi2802

Hi all, I'm Rishitha Yerragogu, about to start my second year in Computer Science and Engineering. I have good experience in C, C++ and OOPS. I did few University mini Projects using C++. I would like to contribute on the project "Space- and time-efficient data format for mass spectrometry data (OpenMS)", I have all the skills and languages needed for the project and am receptive to learn so much more through this. Looking forward to working with you. Thanks! @timosachsenberg @jpfeuffer @poshul

Timo Sachsenberg

@timosachsenberg

Hi @rishi2802 thanks for your interest but GSoC is already running :)

rishi2802

@rishi2802

@timosachsenberg Thanks for your reply, is there any way I can contribute not just for GSoC but for Open Source and learning in General? Also, also could you please suggest me some project that could help me in GSoC'23?

Timo Sachsenberg

@timosachsenberg

hi rishi2802. Sure, we always have plenty to do. You can try to compile OpenMS from source and then take a look at those issues: https://github.com/OpenMS/OpenMS/issues?q=is%3Aissue+is%3Aopen+label%3Afirst-timers-only

rishi2802

@rishi2802

Thanks a lot!

Where communities thrive

People

Repo info

Activity