[BioC] pathview -- ID Problem?

Luo Weijun luo_weijun at yahoo.com
Fri Nov 1 20:04:19 CET 2013


Fabian, 
Sorry for the confusion, I am glad that you figured it out.
Weijun

--------------------------------------------
On Fri, 11/1/13, fabian <fabian at tbi.univie.ac.at> wrote:

 Subject: Re: pathview -- ID Problem?

 Cc: bioconductor at r-project.org
 Date: Friday, November 1, 2013, 12:19 AM

 Hello

 It took me a while but I figured that. Anyway, thank you
 very much for coming back to me for this correction.

 Best
 Fabian

 On 01-11-2013 01:03, Luo Weijun wrote:
 > There was a typo in my original response, id.type
 should be
 > gene.idtype. The correct one is:
 > ..., for you data, you need to specify
 gene.idtype="kegg" with pathview.
 > pv.out <- pathview(gene.data = data_bac_values,
 pathway.id = "05132",
 > species = "sey", gene.idtype ="kegg", out.suffix =
 "test", kegg.native
 > = T, same.layer = F)
 > Weijun
 > 
 > --------------------------------------------

 wrote:
 > 
 >  Subject: Re: pathview -- ID Problem?
 >  To: bioconductor at r-project.org,
 fabian at tbi.univie.ac.at,

 >  Date: Friday, October 18, 2013, 2:53 PM
 > 
 >  Dear Fabian,
 >  You need to update your pathview to the latest
 release
 >  version, which is 1.2.0. It is at:
 >  http://bioconductor.org/packages/release/bioc/html/pathview.html
 >  This current release version includes all recent
 update
 >  features, like multiple samples/time series in
 one graph,
 >  and handling of all KEGG species.Your version is
 >  pathview_1.1.4, which doesn’t really handle
 many minor
 >  KEGG species. To ensure smooth installation, I
 would suggest
 >  you update your R/Bioconductor to the latest
 version too. Of
 >  course, you may always download and install
 pathview_1.2.0
 >  manually from the link above.
 > 
 >  BTW, for you data, you need to specify
 id.type="kegg" with
 >  pathview.
 >  pv.out <- pathview(gene.data =
 data_bac_values,
 >  pathway.id = "05132", species = "sey",
 id.type="kegg",
 >  out.suffix = "test", kegg.native = T, same.layer
 = F)
 >  Read subsection 7.5 Working with species in the
 new vignette
 >  for more details.
 >  Weijun
 > 
 >  --------------------------------------------


 >  wrote:
 > 
 >   Subject: pathview -- ID Problem?
 >   To: bioconductor at r-project.org,
 >  fabian at tbi.univie.ac.at

 >   Date: Friday, October 18, 2013, 1:46
 PM
 > 
 > 
 >   Hello Mr. Weijun Luo
 > 
 >   I started to use pathview to visualize
 my differential
 >  gene
 >   expressin analysis. It is really a
 great help. First I
 >   worked with human data and had no
 problem at all to make
 >  it
 >   run.
 > 
 >   Meanwhile I apply bacterial data and
 do not succeed. I'm
 >  not
 >   sure why. Here is my problem
 > 
 >   When I run the analysis with my input
 > 
 >   head(data_bac_values)
 >   SL1344_0001 SL1344_0002 SL1344_0003
 SL1344_0004
 >  SL1344_0005
 >   SL1344_0006
 >     0.1329739   0.4947647 
 >   -1.4013626   0.5086116 
 -4.0000000 
 >   -0.2028169
 > 
 >   with the following command:
 > 
 >   > pv.out <- pathview(gene.data =
 data_bac_values,
 >   pathway.id = "05132", species = "sey",
 out.suffix =
 >  "test",
 >   kegg.native = T, same.layer = F)
 >   Error in
 mol.data[as.character(items[hit]), ] : subscript
 >   out of bounds
 >   In addition: Warning messages:
 >   1: In node.map(gene.data, node.data,
 node.types =
 >   gene.node.type, node.sum = node.sum)
 :
 >     NAs introduced by coercion
 >   2: In FUN(1:27[[1L]], ...) : NAs
 introduced by coercion
 > 
 >   This is the error thrown out.
 traceback() informs like
 >   this:
 > 
 >   > traceback()
 >   7:
 cbind(mol.data[as.character(items[hit]), ])
 >   6:
 apply(cbind(mol.data[as.character(items[hit]), ]), 2,
 >   node.sum,
 >          na.rm = T)
 >   5: FUN(1:27[[1L]], ...)
 >   4: lapply(X = X, FUN = FUN, ...)
 >   3: sapply(1:n.nodes, function(i) {
 >          kns =
 >   node.data$kegg.names[[i]]
 >          if (node.types[1] ==
 "gene")
 > 
 >              items =
 >   as.numeric(kns)
 >          else items = kns
 >          ord = order(items)
 >          items = items[ord]
 >          kns = kns[ord]
 >          hit = items %in% genes
 >          if (sum(hit) == 0) {
 >          
 >      return(c(kns[1], spacials[i, ],
 rep(NA,
 >   ncol(mol.data))))
 >          }
 >          else if (sum(hit) == 1) {
 >              edata =
 >   mol.data[as.character(items[hit]), ]
 >          
 >      return(c(kns[hit], spacials[i,
 ], edata))
 >          }
 >          else {
 >              node.sum =
 >   eval(as.name(node.sum))
 >              edata =
 >   apply(cbind(mol.data[as.character(items[hit]),
 >              
 >      ]), 2, node.sum, na.rm = T)
 >          
 >      return(c(kns[hit][1],
 spacials[i, ],
 >   edata))
 >          }
 >      })
 >   2: node.map(gene.data, node.data,
 node.types =
 >   gene.node.type, node.sum = node.sum)
 >   1: pathview(gene.data =
 data_bac_values, pathway.id =
 >   "05132", species = "sey",
 >          out.suffix = "test",
 >   kegg.native = T, same.layer = F)
 > 
 > 
 >   I'm aware from the mailing list that
 KEGG is not very
 >   consistant with their ID scheme, which
 causes trouble. I
 >   rechecked, the Id I use (Refseq locus
 tag) is also used by
 >   KEGG (e.g. http://www.genome.jp/dbget-bin/www_bget?sey:SL1344_0001
 >   works fine); I tried it with other IDs
 which I specified
 >  via
 >   gene.idtype and get the error:
 > 
 >   > pv.out <- pathview(gene.data =
 data_bac_values,
 >   gene.idtype = "UNIPROT", pathway.id =
 "05132" ,species =
 >   "sey", out.suffix = "test",
 kegg.native = T, same.layer =
 >   F)
 >   Error in pathview(gene.data =
 data_bac_values, gene.idtype
 >  =
 >   "UNIPROT",  :
 >     No proper gene annotation package
 available!
 > 
 >   This is also truefor other IDs (of
 course I change each
 >  time
 >   the input file accordingly).
 > 
 >   Do you have any idea what causes the
 problem? any hint
 >  would
 >   be very much appreciated. Thank you
 very much in advance.
 > 
 >   best
 >   Fabian
 > 
 >    -- output of sessionInfo():
 > 
 >   > sessionInfo()
 >   R version 3.0.1 (2013-05-16)
 >   Platform: x86_64-redhat-linux-gnu
 (64-bit)
 > 
 >   locale:
 >    [1] LC_CTYPE=en_US.UTF-8   
 >      LC_NUMERIC=C       
 >        
 >    [3] LC_TIME=en_US.UTF-8       
 >   LC_COLLATE=en_US.UTF-8   
 >    [5] LC_MONETARY=en_US.UTF-8   
 >   LC_MESSAGES=en_US.UTF-8   
 >    [7] LC_PAPER=C           
 >        LC_NAME=C     
 >              
 >    [9] LC_ADDRESS=C           
 >      LC_TELEPHONE=C       
 >      
 >   [11] LC_MEASUREMENT=en_US.UTF-8
 LC_IDENTIFICATION=C 
 >        
 > 
 >   attached base packages:
 >   [1] parallel  stats 
 >      graphics  grDevices utils 
 >      datasets  methods 
 >   [8] base     
 > 
 >   other attached packages:
 >    [1] pathview_1.1.4   
 >      org.Hs.eg.db_2.9.0   RSQLite_0.11.4 
 >      
 >    [4] DBI_0.2-7           
 >   AnnotationDbi_1.22.6 Biobase_2.20.1 
    
 >    [7]
 >   BiocGenerics_0.6.0   KEGGgraph_1.16.0 
 >      graph_1.38.3       
 >   [10] XML_3.98-1.1       
 > 
 >   loaded via a namespace (and not
 attached):
 >   [1] grid_3.0.1     
 IRanges_1.18.4 
 >   png_0.1-6       Rgraphviz_2.4.1
 >   [5] stats4_3.0.1    tools_3.0.1 
  
 > 
 > 
 >   --
 >   Sent via the guest posting facility at
 bioconductor.org.



More information about the Bioconductor mailing list