[BioC] KEGGSOAP Question: get.pathways.by.compounds

[Nianhua Li]

Hi, Alan,

I cc'd your question to the list, and please see my answer below:

ALAN SMITH wrote:
> Hi Nianhua,
> Sorry to bother you again.  Your advice helped me get through my
> problems.   I have now run into another problem that is a fairly
> serious problem with get.pathways.by.compounds in KEGGSOAP.  The
> get.pathways.by.compounds only works with your example data.  Change
> it and it does not work.
> 
> pathways <- try(get.pathways.by.compounds(c("cpd:C00033",
> "cpd:C00158"))) #example#
> pathways
> [1] "path:map00020" "path:map00720"
> #It is truncating the pathways only showing the first one in the list#
> bget("LIGAND:C00033")
> bget("LIGAND:C00158")
> 
No, it doesn't. "get.pathways.by.compounds(c("cpd:C00033", "cpd:C00158"))" returns the pathways that
contain BOTH compounds. Try:
   get.pathways.by.compounds(c("cpd:C00033", "cpd:C00033"))
    [1] "path:map00010" "path:map00020" "path:map00430" "path:map00440"
    [5] "path:map00450" "path:map00534" "path:map00620" "path:map00622"
    [9] "path:map00660" "path:map00720" "path:map00920"
   get.pathways.by.compounds(c("cpd:C00158", "cpd:C00158"))
    [1] "path:map00020" "path:map00251" "path:map00252" "path:map00630"
    [5] "path:map00720"
Notice that you always have to give more than one compounds. If there is only one compound, please
repeat it (sorry). This is because of a problem in the function. I will try but probably can't fix
it quickly.
> 
> The get.pathways.by.compounds only works with your example when I try
> this cpd
> bget("LIGAND:C00328")
> I should recover PATH: map00380
> 
> pathways <- try(get.pathways.by.compounds(c("cpd:C00033", "cpd:C00328")))
> pathways
> #character(0) #does not work#
> pathways <- try(get.pathways.by.compounds("cpd:C00328"))
> pathways
> #character(0) #does not work#
Try:
   get.pathways.by.compounds(c("cpd:C00328", "cpd:C00328"))
   get.pathways.by.compounds(c("cpd:C00033", "cpd:C00033"))
Actually get.pathways.by.compounds(c("cpd:C00033", "cpd:C00328")) should return character(0). I
tried it via a java client.
> 
> I have tried many other permutations cpds and none seem to work except
> the example
Sorry for the inconvenience. Hope it works fine from now on...
> 
> Is there anything you can do to fix this problem?  Should I contact
> the people at KEGG and explain the problem?  The main reason I started
> using KEGGSOAP was to get to pathways from neutral masses because the
> important biological interpretation of the data is in the pathways.
> 
> Thanks,
> Alan Smith

best
nianhua