Search Mailing List Archives


Limit search to: Subject & Body Subject Author
Sort by: Reverse Sort
Limit to: All This Week Last Week This Month Last Month
Select Date Range     through    

[go-friends] [go-helpdesk] GO Terms absent from a protein list; negative enrichment?

Sorin Draghici sorin at wayne.edu
Mon Jan 30 16:15:40 PST 2012


Dear Adam,

I am really happy to hear that somebody is paying attention to the
under-represented GO terms. I have been saying for years that this is as
important as the over-representation (see for instance "Use and misuse
of the gene ontology annotations". Nature Reviews Genetics, July 2008, 9
(4) : 509-15 or Chap 24: Uses, misuses and abuses in GO profiling in the
book "Statistics and Data Analysis for Microarrays Using R and
Bioconductor", CRC Press, 2012
http://www.amazon.com/Statistics-Microarrays-Bioconductor-Mathematical-Computational/dp/1439809755/ref=dp_ob_title_bk.


Our tool, Onto-Express, available for free from our lab's web site:

http://vortex.cs.wayne.edu

has been performing tests for both under- and over-enrichment since
2002. Onto-Express also allows you to perform a cut through the GO DAG
at a custom level of abstraction that you define by collapsing and/or
expanding GO nodes. The results can be exported in text files or as
pie-charts, etc.

Be careful, there is another subtlety  here: the NOT annotations. What
is NOT in your set of genes is not the same as things that ARE there but
they are annotated with NOT being associated with a specific GO term.
See the chapter cited above for a full discussion.

Best regards,


Sorin Draghici

Robert J. Sokol, MD Chair in Systems Biology
Professor, Department of Clinical and Translational Science,
Professor, Department of Computer Science
Director of the Intelligent Systems and Bioinformatics Laboratory
Wayne State University
14th Floor, Suite 14200
5057 Woodward,
Detroit, MI 48202
Tel: 313-577-5484
Fax: 313-577-6868
http://vortex.cs.wayne.edu

On 1/30/2012 3:39 AM, Rachael Huntley wrote:
> Hi Adam,
> 
> I'm also forwarding your request to the go-friends list to which many GO
> tool developers are subscribed and might be able to suggest something
> for you.
> 
> Best wishes,
> Rachael.
>>
>> ___________________________________________________
>> From:<adam.byron at manchester.ac.uk>
>> Date: Fri, 27 Jan 2012 06:40:42 -0800
>> To:<go-helpdesk at mailman.stanford.edu>
>> Subject: GO Help query (from website)
>>
>>
>> Email: adam.byron at manchester.ac.uk
>> Name: Tools (from Adam Byron)
>> Text: Hi. I am trying to determine which GO terms are absent from my
>> list of
>> identified proteins. I have performed underrepresentation enrichment
>> analysis
>> in BiNGO and FuncAssociate, but am getting limited information from
>> this. It
>> appears that underrepresentation analysis requires at least evidence
>> for a
>> GO
>> term in the submitted protein list, albeit underrepresented compared
>> to the
>> proteome. Can you recommend any tools or techniques that may be
>> appropriate
>> for determining what is NOT in my dataset, rather than what is?
>> Thanks in advance.
>> Adam
>>
>>
>> _______________________________________________
>> go-helpdesk mailing list
>> go-helpdesk at lists.stanford.edu
>> https://mailman.stanford.edu/mailman/listinfo/go-helpdesk
> 
> 


More information about the go-friends mailing list