#11 After UPDATE; sort pseudogenes

[Midori Harris]

Some pseudogenes have changed status and need
"warning, previously annotated as pseudo"
and GO annotation.
This needs checking after the next GeneDB update.
The current list is attached

from trac ticket:
https://sourceforge.net/apps/trac/pombase/ticket/22

[Valerie Wood]

All pseudogenes need sorting generally, and procedures writing for how they are defined

We should probably remove the pseudogene label from everything except the very degraded, and make sure all the coordinated define the "translation"

[Valerie Wood]

If a gene has a single frameshift, it probably isn;t good to represent them as psuedo genes, so I am making "valid translations" for these, removing the psudo label and flagging as
"
warning, possible frameshift
warning, previously annotated as pseudo
(they may be problems in the sequenced stain and some are problems with the sequence)

This way people will see them in the gene set if they are tring to do comparisons with other strains, or octosporus, japonicus etc. Adding the pseudo flag makes them less visible by excluding them from the protein set....

[Valerie Wood]

More considerations:

Hi Val.

While investigating this I did a Chado query for interactions between
features that aren't gene. One case I found was:
http://thebiogrid.org/275375/summary/schizosaccharomyces-pombe/apn1.html

This is a pseudogene. Does it make sense for it to have interactions
in BioGRID?

It also has GO annotations. That doesn't make sense to me either.

Kim.
Yeah this is a pain,
In this strain there is a stop codon right at the beginning of this sequence and it is non-functional (has NOT annotations)
It is likely that this is not a pseudogene in closely realted stains, and ideally I would represent it as a coding sequence, with no psudo, but with NOT annotations and a note that it is mutatated in 972 h-. Unfortunately tyhere is not easy way to do this (I guess I could force a translation which skipped the stop codon, but this would also look a bit odd.

I will try to come up with a better way of doing this. You could filter the electronic mappings when loading ... but its no big deal. I have a todo list item to go through all of the psuedogenes gradually

val

[Valerie Wood]

Same for this one:
http://thebiogrid.org/280106/summary/schizosaccharomyces-pombe/spcp20c803.html

Kim.

yeah again, this one is a little different , almost certainly not protein coding, as it has no methionine, and is frameshifted.

However, it is possible the deletion of this region could have genetic interactions becasue the trasnscript could have a regulatory role, so it could generate a signal. Or it could be a false positive. I think we just have to live with this, until we decide what to do ...we will discuss pseudogene annotation at a future curation meeting and try to resolve.

Val

[Valerie Wood]

Until somebody moans lowering