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 evidence code usage questions

Zhiyong Lu Zhiyong.Lu at
Wed May 18 12:07:59 PDT 2005

Dear GO friends,

I'd like to share some data we collected about GO annotations with
you, and ask a few related questions of the annotators.  

The lab where I work is looking at the relationships between pubmed
entries (mostly abstracts) and various kinds of annotations.  I looked
at the human gene records in Entrez Gene (as of Jan 19, 2005) to see
which had GO annotations that were associated with a specific PubMed
ID.   Looking at that data by evidence code turned up some interesting
phenomena, and I was wondering of any of the annotators (or anybody
else) had any comments about them. 

Here's the data:

code   w/PMID        %w      w/o      %w/o   total    
IMP       228      98.3%       4      1.7%     232      
ND         68       5.9%    1080     94.1%    1148     
ISS       572      31.1%    1269     68.9%    1841     
TAS     17180      99.9%      22      0.1%   17202    
IC        104      96.3%       4      3.7%     108      
IGI        14     100.0%       0      0.0%      14       
NR          8       0.4%    1858     99.6%    1866     
IEP       175     100.0%       0      0.0%     175      
IEA       872       2.2%   38373     97.8%   39245    
NAS      4660      74.8%    1572     25.2%    6232     
IPI       402      99.3%       3      0.7%     405      
IDA      1959      99.8%       3      0.2%    1962     
total   26242      37.3%   44188     62.7%   70430   

w/PMID means the number (or %) of annotations with that evidence code
which are associated with a specific pubmed id. "w/o" means not
associated with a specific pubmed id.

My questions are:

 1. How can it be that 22 "traceable author statements" are not
    associated with a PMID? 

 2. Why are so many "nontraceable author statements" (nearly 75%)
    associated with a PMID?  What is the association?  Might I expect
    to find a statement justifying the annotation in the article? 

 3. Why are so many (nearly 1000) "inferred by electronic annotation"
    annotations associated with a pubmed id?  What in an article would
    trigger an IEA code rather than some other evidence code? 

Thanks very much!

Zhiyong Lu
Center for Computational Pharmacology
University of Colorado

This message is from the GOFriends moderated mailing list.  A list of public
announcements and discussion of the Gene Ontology (GO) project.
Problems with the list?           E-mail: owner-gofriends at
Subscribing   send   "subscribe"   to   gofriends-request at
Unsubscribing send   "unsubscribe"  to  gofriends-request at

More information about the go-friends mailing list