Unrequired metabolites and reactions in the network?

[niraito]

Hello, I was trying to reproduce pinocembrin production pathways by using:

• pinocembrin as the source,
• parsed (by using RRParser) version of reaction rules (retrorules_rr02_rp2_flat_all.csv of rr02 | 2 to 16 | explicit | RetroPath 2.0) taken from RetroRules website download page, and
• sink file obtained by using rpExtractSink on E.coli(bigg_iAF1260).sbml file (I suppose I have downloaded it from previous version of Galaxy-SynBioCAD)

You can find them collected here: 19-docs.zip

Here is my command using RetroPath2-wrapper on mentioned files, and the result:

python -m retropath2_wrapper 19-docs/E._coli_sink.csv 19-docs/retrorules_parsed.csv 19-repeat-of-11/ --source_file 19-docs/source.csv
/home/nilay/anaconda3/envs/rptools/lib/python3.9/runpy.py:127: RuntimeWarning: 'retropath2_wrapper.__main__' found in sys.modules after import of package 'retropath2_wrapper', but prior to execution of 'retropath2_wrapper.__main__'; this may result in unpredictable behaviour
  warn(RuntimeWarning(msg))
Configuration
 + retropath2_wrapper
     |--version: 2.2.0
 + KNIME
     |--path: /home/nilay/anaconda3/envs/rptools/lib/python3.9/site-packages/retropath2_wrapper/.knime/linux/knime_4.5.0/knime
 + RetroPath2.0 workflow
     |--path: /home/nilay/anaconda3/envs/rptools/lib/python3.9/site-packages/retropath2_wrapper/workflows/RetroPath2.0_r20220104.knwf

Checking input data
   |- InChI
   |- Source in Sink (simple)
Initializing
   |- Formatting files for KNIME
Running KNIME... OK
   |- Checking Source in Sink (advanced)
OK
Results
   |- Checking...
   |--path: 19-repeat-of-11/

When I visualized it on Scope Viewer:

When I zoomed in:

And when I moved just the source molecule:

I see there are many metabolites and reactions. However, the source molecule connects to only a small pathway. If the pathway is just two steps, why are there all these metabolites and reactions?

[tduigou]

Thanks @niraito for pointing out this. We were a bit overwelmed these last days, hence my late reply. I did check the result and it appears there's an issue in the way the scope is computed. I'll see how to correct that in RetroPath2.0. Notice that anyway this does not change the validity of the results, ie the scope is actually a "superset" of what it should be, but enumeration of pathways with rp2paths should not change.