annotating to pseudogenes

[Rama Balakrishnan]

Sue, Suzi, All,

I have been meaning to pipe in earlier but had to deal with a sick  
child. any way..

I was going to raise the same question Sue raised in her email about  
determining functionality.
It is non trivial to determine if a gene is functional or not. We  
(SGD) have been debating this issue for a long time now.
This is one of the main reasons why we have left the unknown  
annotations.

So, I would like to know how do other groups decide what is a  
pseudogene. Do you use some in house analyses to determine if a  
sequence is a pseudogene or do you rely on published literature to  
say that a sequence is a pseudogene based on Blah et al?

Thanks,

Rama


On Mar 14, 2006, at 4:25 PM, Sue Rhee wrote:

> Suzi and all,
>
> Is SO defining sequence elements based not only on the structural/ 
> genome context, but also on functional attributes? If so, I think  
> the definition of SO's pseudogene is perfectly reasonable.  
> Otherwise, I question whether the 'non-functionality' part of the  
> definition for a pseudogene is appropriate in SO.
>
> Even if we stick to SO's definition of a pseudogene, it is not  
> trivial to unequivocally determine that a sequence element is non- 
> functional (unless there are explicit, agreed upon criteria for  
> determining non-functionality that I am not aware of).
>
> It appears that we have two choices here: One is to be more  
> restrictive about calling something a 'pseudogene' in our databases  
> based on a set of agreed upon criteria (e.g. expression, assigned  
> role/function, anything else?). Two is to define a pseudogene  
> without the functionality component and annotate them to unknown  
> unless there is a function/role defined.
>
> These are difficult choices to make, with pros and cons for each.  
> I'm happy with sticking to either choice if we could reach a group  
> consensus. Do you all think this would be possible? How can we best  
> do this?
>
> Thanks,
> Sue
>
>
> Suzanna Lewis wrote:
>> On Mar 14, 2006, at 8:35 AM, Fiona McCarthy wrote:
>>> Hi All,
>>>
>>> I recent found a microarray tiling paper (PMID: 15876366) where the
>>> authors estimated that one fifth of human pseudogenes on chr22 are
>>> potentially transcribed. It seems to me that there must be a  
>>> reason for
>>> this level of transcription, even if we don't know what it is.
>>>
>>> I think it would be reasonable to annotate pseudogenes to GO,  
>>> even if we
>>> have to use function unknown in most cases. Otherwise, we would be
>>> implying that *all* pseudogenes have no function, and this may  
>>> not be the
>>> case.
>> See Peter's reply.
>> But to repeat.
>> All pseudogenes are, by our definition, things that you believe  
>> are truly and sincerely, completely, uttterly dead (shades of the  
>> dead parrot skit).
>> If, in your scientific judgment, they may have some function  
>> (including latent, waiting for recombination, function a la  
>> inactive, cold-storage units) then they are not pseudogenes: in  
>> our shared, common definition.
>> We have agreed that it is useful to have clear definitions, and  
>> that we will use the same terms to mean precisely the same thing.  
>> We must not fall away from this.
>> While we are aware that there are entire communities that use the  
>> term "pseudogene", but mean something different, we must be  
>> consistent ourselves. Which means that we must have other terms  
>> (with the synonym "pseudogene") to describe these different  
>> phenomena.
>> Now, part 2.
>> Given that this thing has a function, then you get into the issue  
>> of what is that function. Once you have agreed that the function  
>> is in there, it is perfectly fine to say "unknown".
>> -S
>>>
>>> As for the SO definition of a pseudogene, I am not sure that I  
>>> would say
>>> that all pseudogenes are non-functional.
>>>
>>> Fiona
>>>
>>>
>>> AgBase Biocurator
>>> Department of Basic Sciences
>>> Box 6100
>>> MS 39762-6100
>>> Mississippi State University
>>> USA
>>> Tel:   (+ 1) 662 325 5859
>>> Fax:  (+ 1) 662 325 1031
>>>
>>> http://www.agbase.msstate.edu/
>>>