[BioC] Pathview with minor species

[Luo Weijun]

I forgot to mention that in pathview vignette there is a dedicated section covering on species related issues, i.e. Section 7.5 “Working with species”:
http://bioconductor.org/packages/release/bioc/vignettes/pathview/inst/doc/pathview.pdf

I just noticed that locus tag is used as the default gene ID for KEGG pathways for B. subtilis (bsu). This is actually true for many bacteria and other minor species (including E coli K12). KEGG does not specify the default gene ID type explicitly, but you may find this out for your species in korg data (the kegg.geneid column).
> data(korg)
> ri=which(korg[,1]=="bsu")
> korg[ri,]
          kegg.code     scientific.name         common.name       entrez.gnodes
              "bsu" "Bacillus subtilis"                  ""                 "0"
        kegg.geneid         ncbi.geneid
         "BSU00020"            "939970"

So you can actually do it in a simple way. You don’t have to map your gene IDs, just specify gene.idtype="kegg" when calling pathview. My analysis with simulated bsu data works well.

bsu.dat.kegg <- sim.mol.data(mol.type="gene",id.type="kegg",species="bsu",nmol=3000)
head(bsu.dat.kegg)
   BSU13160    BSU11080    BSU23500    BSU02510    BSU19920    BSU20590 
-1.15259948  0.46416071  0.72893247  0.41061745 -1.46114720 -0.01890809 
pv.out <- pathview(gene.data = bsu.dat.kegg, gene.idtype="kegg",
pathway.id = "00010", species = "bsu", out.suffix = "bsu.kegg")

A similar example was given on E coli K12 in page 20 of the vignette. I would suggest to go over section 7.5 “Working with species” if you need to work on minor species.


--------------------------------------------


 Subject: Re: Pathview with minor species
 To: bioconductor at r-project.org
 Date: Wednesday, January 29, 2014, 3:25 PM

 Someone asked a similar question on
 using pathview or gage with honey bee. Here is the
 solution:
 http://seqanswers.com/forums/showthread.php?t=35472#6

 Like honey bee, B. subtilis (bsu) is not a major species
 with Bioconductor gene annotation package as those listed in
 data(bods) under pathview package. similarly you need to map
 Locus_tags to Entrez gene ID manually. Again, you can
 download the gene_info data file (for all bacteria) from
 NCBI ftp site:
 ftp://ftp.ncbi.nlm.nih.gov/gene/DATA/GENE_INFO/Archaea_Bacteria/Bacteria.gene_info.gz
 Column 2-6 should be (Entrez) GeneID, Symbol, LocusTag,
 Synonyms, dbXrefs. Then follow the route described in the
 honey bee case.

 HTHs.

 --------------------------------------------
 On Wed, 1/29/14, David wrote:

  Subject: Pathview with minor species

  Date: Wednesday, January 29, 2014, 11:09 AM
  
  Dear Mr Luo,
  
  In our course we got B. subtilis (bsu) microarray data
  and I want to visualize some expression changes with
  Pathview. But I have a problem. We got Locus_tags like
  BSU00010 and no gene IDs and I don't know whether Pathview
  can use it or not. I really don't understand the part
 about
  it in the vignette. So I thought of converting them to
  UniProt IDs. But first I wanted to do a test run. So I
  searched for the UniProt IDs of 3 genes of the glycolysis
  from B. subtilis and put them into a data.frame with
  fictional data (dummy).
  
  library(pathview)
  dummy<-data.frame(name=c("P18159","P80860","P13243"),ratio=c(1,0,-1),row.names="name")
  dummy<-as.matrix(dummy)
  
  pv.out <- pathview(gene.data = dummy[, "ratio"],
  pathway.id = "00010",species = "bsu", out.suffix =
 "dummy",
  kegg.native=T, gene.idtype="UNIPROT")
  
  
  When I tried it this way, an error occured: No proper gene
[[elided Yahoo spam]]
  Afterwards I tested it with 3 genes from human and hsa as
  species and than it worked fine. Nothing I tried to solve
  the problem, worked. I have the latest version of
  Bioconductor and Pathview installed. Does Pathview not
 work
  with Bacillus subtilis subsp. subtilis 168 (=bsu) even
  though it's a KEGG organism? Or what am I doing wrong?
  
  I look forward to hearing from you.
  
  Sincerely yours,
  David