[Annotation] [Go] Usage of the With/From Column for IEA ?

Tue Jun 24 13:31:11 PDT 2008

Hi,

I personally don't have a problem with dropping the requirement for the 
with column for IEAs to be comparable to the rules for ISS, though I am 
certainly not speaking for all of SGD.

However, having tried to come up with a system to store method names in 
the with column after the decision at the Jan 07 meeting that we would do 
this, I think this is a bad idea. It is non-trivial to identify the name 
of a method from many papers, and unless we are going to start keeping 
some sort of file to control what goes in there, these method names could 
become a muddle of free text in the with/from column.

So between the two options you suggest, I think only the first is 
feasible.

But, I think the best idea is what Kara suggested, that whether or not the 
annotation is manual vs electronic should be recorded separately. Then the 
evidence codes can be used as appropriate for either manual or electronic, 
following the appropriate rules for the with column for that evidence 
code. I do realize that this would be a big change and won't happen 
quickly though...

-Karen

On Mon, 23 Jun 2008, Gwinn Giglio, Michelle wrote:

> Hi all,
>
> I think there are two options:
>
> 1.  We extend the same rules that currently apply to ISS to IEA.  Namely,
> that when tools such as tRNAscan and snoSCAN are used there does not need to
> be an entry in "with/from".
> 
> OR
>
> 2.  We come  up with a  system to store this information in "with/from"
> (which would also apply to ISS).
> Karen proposed such a system way back when this discussion for ISS was first
> underway and her system would have worked well I think.
>
> Votes?
>
>
> Michelle
>
>
>
>
> On 6/21/08 10:33 PM, "Pankaj Jaiswal" <pj37 at cornell.edu> wrote:
>
>> Can we have a resolution on this? We are waiting to checkin the
>> association file.
>>
>> Pankaj
>>
>> Gwinn Giglio, Michelle wrote:
>>>
>>> Ben,
>>>
>>> No one is disputing the importance of the "with" fields and I think we all
>>> know why we have them.  However, even making the "with" field always
>>> mandatory will not make people annotate correctly.  It is easy to get around
>>> the spirit of these "rules" if someone wants to.
>>>
>>> The bottom line is we have had this debate more than once and have reached
>>> the decision not to require "with" for all ISS annotations - it only  makes
>>> sense that the IEAs using the same kinds of information would be treated the
>>> same way.
>>>
>>> Let's think about the people who are trying to do a thoughtful and careful
>>> job of annotating using best GO practices.  They will read the guidelines
>>> and they will try to follow them.
>>> As for those who use bad annotation practices, they will produce bad
>>> annotation whether or not we make them populate "with".
>>>
>>> Putting the methods/tools like tRNAscan and snoRNA scan into "with" would be
>>> mixing methods and accessions-of-items-found-with-methods in the "with"
>>> column.  Many GO folks have expressed that they don't want to do this.  As
>>> Pankaj and others have pointed out, there is always the option of yet
>>> another column for this.  Or, others have argued it can be captured in the
>>> reference field.
>>>
>>> Anyway, we have a policy in place.  Of course we've changed our minds in the
>>> past before.....
>>>
>>> Michelle
>>>
>>>
>>>
>>>
>>>
>>> On 6/12/08 2:38 PM, "Benjamin Hitz" <hitz at genome.stanford.edu> wrote:
>>>
>>>> WITH fields are very important for IEA and "true" ISS where you are
>>>> inferring an annotation based on sequence similarity.
>>>> The reason is to prevent circular associations via sequence similarity.
>>>>
>>>> That is why the WITH column was made mandatory for IEAs and (old)
>>>> ISS.  From a database/machine loading perspective, the only way to
>>>> enforce this is to make the with field MANDATORY.
>>>>
>>>> Ben
>>>>
>>>> On Jun 12, 2008, at 10:02 AM, Gwinn Giglio, Michelle wrote:
>>>>
>>>>>
>>>>> I think we do need some kind of consensus - anyone else want to
>>>>> voice an
>>>>> opinion?
>>>>>
>>>>> Michelel
>>>>>
>>>>>
>>>>>
>>>>> On 6/12/08 11:48 AM, "Pankaj Jaiswal" <pj37 at cornell.edu> wrote:
>>>>>
>>>>>> Great. Do we need a formal word or voting on this request. We are
>>>>>> waiting to submit some of our annotations that didn't make it through
>>>>>> due to the mandatory 'with' rule.
>>>>>>
>>>>>> Pankaj
>>>>>>
>>>>>> Gwinn Giglio, Michelle wrote:
>>>>>>> Hi,
>>>>>>>
>>>>>>> Well, considering our new policy for ISS not requiring "with" for
>>>>>>> the very
>>>>>>> types of evidence you mention, it seems IEA should not require
>>>>>>> "with" as
>>>>>>> well......
>>>>>>>
>>>>>>> Michelle
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> On 6/11/08 1:12 PM, "Pankaj Jaiswal" <pj37 at cornell.edu> wrote:
>>>>>>>
>>>>>>>> I didn't hear from the group on this. Could you please help.
>>>>>>>>
>>>>>>>> Pankaj
>>>>>>>>
>>>>>>>> Pankaj Jaiswal wrote:
>>>>>>>>> http://www.geneontology.org/GO.evidence.shtml#iea
>>>>>>>>>
>>>>>>>>> Ref: From the above site
>>>>>>>>>
>>>>>>>>> Usage of the With/From Column for IEA
>>>>>>>>>
>>>>>>>>> At the January 2007 GOC meeting, it was agreed that it will be
>>>>>>>>> required
>>>>>>>>> to make an entry in the with/from column for all annotations
>>>>>>>>> made after
>>>>>>>>> May 1, 2007 when using this evidence code to indicate what
>>>>>>>>> individual
>>>>>>>>> sequences, sequence objects, methods, keyword mapping files,
>>>>>>>>> etc. are
>>>>>>>>> the basis of the annotation. When multiple entries are placed in
>>>>>>>>> the
>>>>>>>>> with/from field, they are separated by pipes.
>>>>>>>>>
>>>>>>>>> ------------
>>>>>>>>> Based on this rule, would you please suggest, what would be the
>>>>>>>>> value
>>>>>>>>> for the 'WITH' column.
>>>>>>>>>
>>>>>>>>> - if we are making a TMHMM prediction for a 'putative
>>>>>>>>> tansmembrane'
>>>>>>>>> annotation
>>>>>>>>> - if we have used TargetP/SignalP/Predotar/Psort to predict a
>>>>>>>>> putative
>>>>>>>>> cellular localization
>>>>>>>>>
>>>>>>>>> None of the above prediction softwares give appropriate value to
>>>>>>>>> be
>>>>>>>>> filled in the 'WITH' column.
>>>>>>>>>
>>>>>>>>> If you agree our suggestion is to relax this rule for IEA's.
>>>>>>>>>
>>>>>>>>> Pankaj
>>>>>>>>>
>>>>>>>> _______________________________________________
>>>>>>>> Go mailing list
>>>>>>>> Go at geneontology.org
>>>>>>>> http://fafner.stanford.edu/mailman/listinfo/go
>>>>>>>
>>>>> _______________________________________________
>>>>> Go mailing list
>>>>> Go at geneontology.org
>>>>> http://fafner.stanford.edu/mailman/listinfo/go
>>>> --
>>>> Ben Hitz
>>>> Senior Scientific Programmer ** Saccharomyces Genome Database ** GO
>>>> Consortium
>>>> Stanford University ** hitz at genome.stanford.edu
>>>>
>>>>
>>>>
>>>> _______________________________________________
>>>> Annotation mailing list
>>>> Annotation at geneontology.org
>>>> http://fafner.stanford.edu/mailman/listinfo/annotation
>>>
>>> _______________________________________________
>>> Go mailing list
>>> Go at geneontology.org
>>> http://fafner.stanford.edu/mailman/listinfo/go
>>>
>
> _______________________________________________
> Go mailing list
> Go at geneontology.org
> http://fafner.stanford.edu/mailman/listinfo/go
>