[BioC] How does pathview deal with enzyme with several coding genes?

[Luo Weijun]

This is not an error message. Just a friendly reminder that nothing (genes or compounds) can be mapped to the specified pathway. In your case, this is because there is only 1 gene in your data.


--------------------------------------------
On Tue, 2/18/14, 刘鹏飞 <liupfskygre at gmail.com> wrote:

 Subject: Re: How does pathview deal with enzyme with several coding genes?

 Cc: "bioconductor at r-project.org list" <Bioconductor at r-project.org>
 Date: Tuesday, February 18, 2014, 7:52 AM

 Thank you for your
 comprehensive explanation，I read some parts of the
 vignette in details as you mentioned, realized several ways
 to visualize my data. Pathview is really a nice
 tools！ 

 I also found a little problem, although it does not
 affect my use.when the list of expression data
 has only one gene, the pathview would not map the
 data to the kegg map and send error message:
 No of the genes or compounds mapped to the
 pathway!Argument gene.idtype or cpd.idtype may be
 wrong.Working in directory
 C:/Users/microbe/DocumentsWriting image file
 pth00480.pthmeanGID.png

 I do not whether other user has similar
 problem. 
 Regards!








 As described in Page 7 of pathview vignette:

 “..Note in native KEGG view, a gene node may represent
 multiple genes/proteins with similar or redundant functional
 role. The number of member genes range from 1 up to several
 tens. They are intentionally put together as a single node
 on pathway graphs for better clarity and readability.
 Therefore, we do not split node and mark each member genes
 separately by default. But rather we visualize the node-wise
 data by summarize gene-wise data, users may specify the
 summarization method using node.sum arguement.”


 Native KEGG graph use the most informative name to label
 multiple-gene nodes. When Graphviz view or 2-layer graph is
 specified, we choose to use the most common gene symbol
 (among multiple ones).



 Check the function documentation of pathview:

 ?pathview

 node.sum argument, character, the method name to calculate
 node summary given that multiple genes

 or compounds are mapped to it. Potential options include
 "sum","mean",
 "median","max", "max.abs" and
 "random". Default node.sum="sum".



 Below are the data you provided. You may want to check
 whether the fold change sum of your 3 genes is about 0.299
 or 0.4.

 > Gene  
  Mtc_1383/aroF; Mtc_2501/fbaA; Mtc_0384/fbaB;

 > Log2fc  -1.22;-0.118;1.645;



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

 On Mon, 2/17/14, 刘鹏飞 <liupfskygre at gmail.com>
 wrote:



  Subject: Re: How does pathview deal with enzyme with
 several coding genes?

  To: "bioconductor at r-project.org
 list" <Bioconductor at r-project.org>



  Date: Monday, February 17, 2014, 2:33 AM



  Hi,I found in the

  mol.data that pathview just
 do some add calculation, for

  example, acs-1=-1.7283407; acs-2=-4.1076455, in the

  mol.data, just acs-1 was chosen, and data was set as

  acs-1=-5.835986205; but,
 for Gene   

  Mtc_1383/aroF;

  Mtc_2501/fbaA;

  Mtc_0384/fbaB;

  Log2fc  -1.22;-0.118;1.645;

  in mol.data, fbaB=0.299; but

  aroF+fbaB=0.4, which is not equal to 0.299;

  So, I am confused on this issue

  now!and more, if several genes

  encoding for different subunits
 of a enzyme,  all get a log2

  fold was less than 1, but add all them up would exceed
 1,

  and give us a wrong information. or like the case
 above,

  which also give us some wrong information.

  why the data in the mol.data is

  not the average of several genes encoding for
 different

  subunit of the same enzyme? And for iso-functional
 enzyme,

  could we display them separately(like fbaB and aroF)?





  Thanks!





  2014-02-17 14:37 GMT+08:00 刘鹏飞 <liupfskygre at gmail.com>:



  Dear all, Now I am using

  pathview to map my RNAseq expression data to the
 keggmap of

  my organism ‘mez’. I want to know how
 pathview

  deal with the colour of enzyme with several conding
 genes,

  for example:



  In pathway: Glycolysis / Gluconeogenesis, mez00010,
 the

  gene node (enzyme 4.1.2.13) has 3 genes related to
 it.

  Gene   

  Mtc_1383/aroF; Mtc_2501/fbaA; Mtc_0384/fbaB;



  Log2fc 

  -1.22;-0.118;1.645; aroF

  and fbaB all significantly regulated and has a |log2
 fold

  change| >1.  Another example

  is that one enzyme has several subunits, each has a
 encoding

  gene, but with different expression level(here I mean
 the

  log2 fold change of treat/ck).



   I found pathview just choose to plot

  one data onto the map(eg, fbaB and aroF, it just
 choose

  fbaB, due to it appear first in the gene.names?).
 which one

  will pathview choose to plot? How could I display all

  information on one plot, like aroF and fbaB, separate
 and

  display them on the map? what about genes encoding
 for

  different subunits of a gene?



   Thanks!

  --

  Pengfei Liu, PhD Candidate



  Lab of Molecular Ecology -

  Max Planck Partner Group



  College of Resources and Environmental Sciences

  China Agricultural University

  No.2 Yuanmingyuanxilu, Beijing, 100193

  P.R. China



  Tel: +86-10-62731358

  Fax: +86-10-62731016

   

  E-mail: liupfskygre at gmail.com

















  --

  Pengfei Liu, PhD Candidate

  Lab of Molecular Ecology -

  Max Planck Partner Group



  College of Resources and Environmental Sciences

  China Agricultural University

  No.2 Yuanmingyuanxilu, Beijing, 100193

  P.R. China



  Tel: +86-10-62731358

  Fax: +86-10-62731016

   

  E-mail: liupfskygre at gmail.com














 -- 
 Pengfei Liu, PhD Candidate
 Lab of Molecular Ecology -
 Max Planck Partner Group

 College of Resources and Environmental Sciences
 China Agricultural University
 No.2 Yuanmingyuanxilu, Beijing, 100193
 P.R. China

 Tel: +86-10-62731358
 Fax: +86-10-62731016
  
 E-mail: liupfskygre at gmail.com