[Go] question about download page

[Chris Mungall]

We should be using ECO here rather than hardcoding ad-hoc lists of  
codes in various pieces of code

There should be some explicitly declared property in ECO that RCA and  
IEA share - either via shared is_a parent or via a relationship to  
some aspect of the type of evidence.

At the moment we have:

   is_a ECO:0000043 ! inferred from in-silico analysis
    is_a ECO:00000067 ! inferred from electronic annotation ***   
[SYNONYM: "IEA" (related)]
    is_a ECO:0000033 ! traceable author statement [SYNONYM:  
"TAS" (related)]
    is_a ECO:0000053 ! inferred from reviewed computational analysis  
***  [SYNONYM: "RCA" (related)]

Which seems dubious. If TAS is placed correctly then we can perhaps  
use inferred from in-silico analysis. But I think ECO needs further re- 
arrangement or changes such that we have an evidence type "inferred  
SOLELY from in-silico analysis"

On Apr 16, 2009, at 8:01 AM, Valerie Wood wrote:

>
>
> On the download page, the number of annotations for each organism  
> are reported and the number of these which are IEA are specified.
> Should this also specify the number which are non-RCA (as these are  
> generally function predictions, with less support than the IEA  
> mappings, and users should be made as aware of the 'dangers' of  
> these,  as of the IEA data).
>
> Val
>
>
> -- 
> The Wellcome Trust Sanger Institute is operated by Genome Research  
> Limited, a charity registered in England with number 1021457 and a  
> company registered in England with number 2742969, whose registered  
> office is 215 Euston Road, London, NW1 2BE.  
> _______________________________________________
> Go mailing list
> Go at geneontology.org
> http://fafner.stanford.edu/mailman/listinfo/go
>