ccms-ucsd / gnps_workflows Goto Github PK

Hi Ming, If possible, would be useful to be able to highlight a node by either m/z or cluster index in the browser,

Thank you

Where we generate the text programmatically and show the text like in feature networking

The issue was already reported to @alexeigurevich and Hosein, and is caused by the MGF parser, more specifically the 'ION" line in the MGF from MZmine:
ION=[M+H]+
with "[", "]", and "+" that are special characters.
The issue is observed both for Dereplicator and Dereplicator+
https://gnps.ucsd.edu/ProteoSAFe/status.jsp?task=4093ac6cf5d848ca9f166673b3fb496c
https://gnps.ucsd.edu/ProteoSAFe/status.jsp?task=6651b66b49d744788cc3e11640c7566f

The work around for now is to delete these lines in the MGF file.

We were discussing having a linkout button from the MolNet V2 results page to to Dereplicator (VarQuest) workflow.
This would pre-select the clustered MGF file in the Dereplicator workflow.

The FEATURE-BASED-MOLECULAR-NETWORKING workflow version 1.2.3 links out to MS2LDA_MOTIFDB version 1.1.1 (Analyze with MS2LDA):

https://gnps.ucsd.edu/ProteoSAFe/status.jsp?task=b817262cb6114e7295fee4f73b22a3ad

The .graphml output looks fine, however, the 'View Motif Labels for Spectra' table for MS2LDA_MOTIFDB version 1.1.1 is empty:

https://gnps.ucsd.edu/ProteoSAFe/result.jsp?task=c03bb8a8f5e74cb89a0b30b815f841b5&view=view_motifs

Using the notebook the correct scores are displayed in the resulting graphml file; when running GNPS-MolNetEnhancer, the scores are not present in the graphml file for classified nodes.

The "View External Ressources" doesn't seem to work anymore.
https://gnps.ucsd.edu/ProteoSAFe/gnpslibraryspectrum.jsp?SpectrumID=CCMSLIB00005436490#%7B%7D

Link on opening page redirects to non-existing workflow on proteomics2 - let's correct to the correct link (https://gnps.ucsd.edu/ProteoSAFe/index.jsp?params=%7B%22workflow%22:%22MOLNETENHANCER%22%7D)

In the FBMN workflow, adding a button "Add a spectrum" for direct addition to GNPS library in the network visualizer.

Create MASST Logo

Having a download button in the results interface to the MGF, feature table and metadata.
Also discussed making available the "converted feature" table generated by the converting script.

unknown error message in FBMN workflow; possibly due to IIN part of feature finding in mzMINE?

https://gnps.ucsd.edu/ProteoSAFe/status.jsp?task=a17860355635486e96116f7555e2f008

FAILED[Clone][Restart][Delete]Traceback (most recent call last): File "/data/ccms-gnps/tools/feature-based-molecular-networking/scripts/convert_networks_to_graphml.py", line 31, in main() File "/data/ccms-gnps/tools/feature-based-molecular-networking/scripts/convert_networks_to_graphml.py", line 26, in main nx.write_graphml(G, sys.argv[5], infer_numeric_types=True) File "</data/ccms-gnps/tools/anaconda3/lib/python3.5/site-packages/decorator.py:decorator-gen-620>", line 2, in write_graphml_lxml File "/data/ccms-gnps/tools/anaconda3/lib/python3.5/site-packages/networkx/utils/decorators.py", line 240, in _open_file result = func_to_be_decorated(*new_args, **kwargs) File "/data/ccms-gnps/tools/anaconda3/lib/python3.5/site-packages/networkx/readwrite/graphml.py", line 150, in write_graphml_lxml writer.dump() File "/data/ccms-gnps/tools/anaconda3/lib/python3.5/site-packages/networkx/readwrite/graphml.py", line 702, in dump self._xml_base.exit(None, None, None) File "src/lxml/serializer.pxi", line 716, in lxml.etree.xmlfile.exit (src/lxml/lxml.etree.c:137075) File "src/lxml/serializer.pxi", line 993, in lxml.etree._IncrementalFileWriter._close (src/lxml/lxml.etree.c:141610) File "src/lxml/serializer.pxi", line 999, in lxml.etree._IncrementalFileWriter._handle_error (src/lxml/lxml.etree.c:141711) File "src/lxml/serializer.pxi", line 195, in lxml.etree._raiseSerialisationError (src/lxml/lxml.etree.c:131087) lxml.etree.SerialisationError: unknown error -1481614473 Tool execution terminates abnormally with exit code [1]

TODO: find link

The job finishes and I get the coloration. Is there a documentation page for this yet? It is not readily apparent which functional classes the colors refer to.

https://gnps.ucsd.edu/ProteoSAFe/status.jsp?task=d9c4449a61d645fca21ef3761e75a058

Analog matches are not reported in the graphml nor teh cytoscape output.
Woudl also be great to get them wit unique column header ( eg XXX_analg) to distuingish from exact library hits.
Thanks!

Utilize the charge to better calculate the parentmass:

https://gnps.ucsd.edu/ProteoSAFe/status.jsp?task=0aa2f6a6463c404ba860a75cce7a17b0

parent mass should be precursor mz * charge - (charge - 1)

https://gnps.ucsd.edu/ProteoSAFe/status.jsp?task=bec183daa7b946bab1ed2d51f7c18d3b

I tried without Normalisation and it seems fine, so the issue is there.
https://gnps.ucsd.edu/ProteoSAFe/status.jsp?task=0061cbb3f9224de183e3c109d8f2d125

Updating the description to clarify the issue with 30 days deletion (including when account in provided).
I am not sure where is the code for that.

• MSHub GC
• MSHub Library Search/Networking
• Merge Polarity Networks
• Mol Net Enhancer
• MS2LDA
• MASST

Downloading edge data from:

https://gnps.ucsd.edu/ProteoSAFe/result.jsp?task=0aa2f6a6463c404ba860a75cce7a17b0&view=download_cytoscape_data

This link includes edges, but these edges do not include self loops in the tsv file in the folder:

networking_pairs_results_file_filtered

Working with one molecule - https://proteomics2.ucsd.edu/ProteoSAFe/result.jsp?task=880196ae2378407ab95f9c5ba5a86397&view=summary_view
Working-ish with mgf from mzmine2 - https://proteomics2.ucsd.edu/ProteoSAFe/status.jsp?task=2901c3583fbe479ebb9a83feb568642c
Broken with mzXML - https://proteomics2.ucsd.edu/ProteoSAFe/status.jsp?task=9e7c726606ea4b96a688310c9ba65c0e

The precursor mass is not a column in the downloaded graphml file for MS2LDA workflow. Please consider adding it. Thank you!

Submitted by @chasemc

There is no way to add directly from MAAST results

MS-DIAL doesn’t output the filename extension (11111_P4_RB1_01_428.mzML) but just (11111_P4_RB1_01_428). After conversion with the "formater" it gives "11111_P4_RB1_01_428 Peak area".
Thus, the user should provide in the metadata table for FTBM job with MS-DIAL the file name without extension in the “sample name" column. @KyobinKang

@mwang87 Reminder for the " Peak area" as general tag for the quant column.

The issue happends for a molecular network job derived from a MAAST job using the function "Analyze with Molecular Networking". When launching the job, some input files selected are causing "Could not find file for f.MSV000079015...]

MAAST: https://gnps.ucsd.edu/ProteoSAFe/status.jsp?task=6179f662018f479dbdccf07c876e6076

As a result, the job cannot be launched.

Delta mass from ion identity networks input are not available in graphml output
This infoprmation is found in the GNPS_edge_annotation table column header; "Annotation" from string: " [M+NH4]+[M+Na]+dm/z=4.955" in this case 4.955

https://gnps.ucsd.edu/ProteoSAFe/status.jsp?task=fc9f0115439a4748a823c28c1267eafc

The above jobs fails due to a missing (VarQuest) file - but in a notebook with the provided task ids it runs fine. I wonder why....

Another job with a DEREPLICATOR task ID did run okay: https://gnps.ucsd.edu/ProteoSAFe/status.jsp?task=69ca43cd5f87463d8ea605399f73afca

https://gnps.ucsd.edu/ProteoSAFe/status.jsp?task=0bee914196074008860683330fd6df3c

suggested enhancement to CCMS library entries - add IUPAC field

Updating the documentation links on the in silico page

Interested in using the GNPS .qza output to run the qiime2 diversity plugin. I want to run the analysis on a rowsum normalized table but would need to make the final table in ppm in order for QIIME2 to accept it.
Would there be an plugin or automated way to create this ppm table? Would it maybe be beneficial to also have any other normalization options?

Should handle cases:

1. Where the is not intensity, only masses
2. Too few peaks given query parameters
3. Extra random characters that can't be parsed

Example inputs:

305.0830, 3000

144.0805

I was reading the preprint by Allard et, and indeed it would be very valuable in the long run to provide a one or multiple species name with the GNPS library entries. Not free text like tags, but rather something similar to the ontology in the MassIVE submission page.
https://www.biorxiv.org/content/biorxiv/early/2019/07/14/702308.full.pdf
See the discussion after line 440.