Search Mailing List Archives
Draft evidence code ontology
midori at ebi.ac.uk
Fri Feb 27 06:21:43 PST 2004
> 4. Becky: 'There are several evidence codes under 'inferred from direct assay',
> which I think should not be, and should be
> moved up a level.
> inferred from genetic interaction
> inferred from mutant phenotype
> inferred sequence similarity
> inferred from expression pattern'
> Midori made the same point.
> .. I disagree, these are all direct assays
> 6: Becky: 6. I think it may be nice to have a general 'inferred from experiment'
> evidence code for all the IDA, IPI, IGI, IEP
> and IMPs?
> .. I think IDA _is_ inferred from experiment
... there is an important distinction to make. In the context of
annotating GO terms to gene products, 'direct assay' is used -- by all
curators doing GO annotation, as far as I am aware -- in a much more
restricted sense than 'inferred from experiment'. For a GO annotation to
cite IDA, the evidence must not only be an experiment (as opposed to a
computation), but the experiment must directly detect what the GO term
For example, a mutant may show reduced levels of protein synthesis, in an
experiment that measures synthesis directly (e.g. by 35S-Met
incorporation). The phenotype nevertheless provides indirect, not direct,
evidence that the product of the mutated gene is directly involved in
In a more general context, there are many possible applications of
evidence codes for which a broader interpretation of 'direct assay' is
entirely reasonable. In the example above, it would make perfect sense to
say that protein synthesis was assayed directly.
But I think it is very important for GO annotations to maintain the
distinction direct observation of something versus direct assay for the
function, process, or component named by the GO term. It is indispensable
if we are to use the evidence codes wintin GO as originally intended -- as
a means of evaluating the reliability of a given GO annotation. How to do
so, whilst still allowing the evidence ontology to be used in contexts
other than GO annotation, is one of the challenges of developing the
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 geneontology.org
Subscribing send "subscribe" to gofriends-request at geneontology.org
Unsubscribing send "unsubscribe" to gofriends-request at geneontology.org
More information about the go-friends