#451 allele descriptions/synonyms/ memory

[Valerie Wood]

Because there are very interrelated i have combined them into a single ticket

1. Reuse allele info (name and description)
   We should be able to reuse alleles for a specific gene if we have used an allele in a previous
   session, (i.e they should appear in the autocomplete)
AND if we add an allele name, the
   description could be populated automatically (currently this happens within a session but
   not between sessions).

I think this was partially done, because in this (now closed, duplicate) item it worked partially int he test tool, but not in the main tool
https://sourceforge.net/p/pombase/curation-tool/388/
The way it works in the test tool, is nearly right but not quite. If I type cdc2, I also see
alleles previously curated for cdc22 cdc20, cdc25 etc, so this should be restricted to the gene
being annotated.

1. we can probably also assume that allele names are case insensitive.
   (e.g.from cdc2) cdc2-3w and cdc2-3W

2. Report any occurrences where the allele descriptions disagree with a warning

"This allele has previously been described as xxxx, please contact curators if your
allele has a different description"

Exception 1: if one of the descriptions is "unknown" or "other" in which case an informative
description can replace.
example for case (from cdc2) cdc2-3w (unknown) and cdc2-3W (C67Y)

Exception 2: We might sometimes describe an allele as its amino acid sequence and sometimes
the nucleotide sequence (curators have decided that if we have both available we will capture
nucleotide because it is more informative. However, in different papers we may be provided only with the AA change, so it is possible we will capture both in which case we would like to store both.

It should be possible to figure out if the nucleotide and protein change reported are equivalent (for example A2F (amino acid) = x4|5|6y nucleotide and only report a warning if they disagree.
(in PomBase we will show amino acid change, but the nucleotide will show when you mouse over…
add to jira tracker)

LATER check for example that amino acid 2 (above) is really an "A", (because sometimes sequence coordinates will have changed since alleles were published) y we could do something with implementing checks like this as part of a student task maybe?…. this is a low priority part of this ticket for now so if this is tricky, take this item out and put it in a new ticket for later (I just enacted to flag it here for completion)

1. if User A enters allele
name ade10-102
description K123X
   User B enters 
name ade10-K123X
description K123X
   These clearly refer to the same allele so behind the scenes the first occurance will be 

   allele_name, or allele_primary_name
 all subsequenct names will become synonyms.

If it is too complicated to work out the precedence based on the PubMed ID just do them in the order the database knows about them. We can fix manually if a user complains
(I do not envisage that they will be very fussy about this).
This should ensure that the allele name and description is unique (per gene at least).
That means we can auto-fill the name if the user types the description first.

[Kim Rutherford]

If I type cdc2, I also see alleles previously curated for cdc22 cdc20, cdc25 etc, so this should be restricted to the gene being annotated.

That's fixed now.

There are quite a lot of jobs to do for this ticket. Is it all priority 7 stuff?

[Kim Rutherford]

If I type cdc2, I also see alleles previously curated for cdc22 cdc20, cdc25 etc, so this should be restricted to the gene being annotated.

That's fixed now.

The allele lookup was disabled in the main tool. It's really fixed now. Sorry!

[Antonia Lock]

This is great!

[Valerie Wood]

probably now. Could you transfer anything outstanding to new tickets with priority 5. We will bump up priority if necessary

Val

[Kim Rutherford]

I've created separate tickets for the bits that still need doing.