#63 updating annotations/merging genes

[Tanya Berardini]

Please preserve the entered_by and date_entered
information when creating a new annotation after
updating an existing annotation.

When I merge two gene models and the old gene model had
annotations by someone else, the annotations move over
but are now in my name with a date of today. This also
happens when I update an existing annotation.

See me for more details if necessary.

Thanks,

Tanya

[Danny Yoo]

Logged In: YES
user_id=49843

What about the date_annotated? Should that be preserved as
well? For the moment, I'll assume that date_annotated
should also be preserved across changes.

I've implemented the change on my test instance here:

http://shoebox.stanford.edu:8080/pub

Please review and see that it looks ok to you. Hope this helps!

[Tanya Berardini]

Logged In: YES
user_id=579762

Hi Danny,

I think we're on our way but it's not quite behaving the way
I expect. I've outlined the expected behavior below, please
let me know if you have any questions:

When updating an existing annotation:

entered_by Suparna
entered_date 2004-09-09
updated_by Sue
updated_date 2005-09-05

Expected result:

entered_by Suparna (same)
entered_date 2004-09-09 (same)
updated_by Tanya (new)
updated_date 2005-09-28 (new)

Rationale: I have revisited the paper, I want to tag this
as being updated by me. I also want to preserve the
identity of the original annotator so that I can consult
that person about this annotation, should I need to.
For merging two gene models: Annotation linked to gene
model that is to be obsoleted:

entered_by Suparna
entered_date 2004-09-09
updated_by Sue
updated_date 2005-09-05

New annotation to retained gene model should look exactly
like the old one with respect to entered and updated by/date
information.

entered_by Suparna (same)
entered_date 2004-09-09 (same)
updated_by Sue (same)
updated_date 2005-09-05 (same)

I believe the current behavior is like this, for both
merging and updating:

entered_by Tanya (new and unexpected)
entered_date 2004-09-09 (same and good)
updated_by Tanya (new and good)
updated_date 2005-09-05 (same and unexpected)

Again, I know this is a little confusing so please come by
and ask if you need clarification.

Thank you!

Tanya

Rationale: I have not revisited the annotations at all, all
I've done is merge gene models. Since the
date_last_synchronized for these annotations should be null,
they should still be eligible for the pubtoTAIR pipeline.

[Tanya Berardini]

Logged In: YES
user_id=579762

Whoops. Should have moved the rationale for the gene merging
solution up. Hope that wasn't too confusing.

[Danny Yoo]

Logged In: YES
user_id=49843

Ok, makes sense. Will implement this tomorrow.

[Danny Yoo]

Logged In: YES
user_id=49843

Fix implemented; can you check? Here's the test url:

http://tesuque.stanford.edu:9999/pubtest

The entered_by and date_entered fields should be preserved now.

[Tanya Berardini]

Logged In: YES
user_id=579762

Hi Danny,

It's half way right. The result of updating an existing
annotation is correct.

Before:

entered_by Suparna
entered_date 2004-09-09
updated_by Sue
updated_date 2005-09-05

After:

entered_by Suparna (same)
entered_date 2004-09-09 (same)
updated_by Tanya (new)
updated_date 2005-09-28 (new)

But the result of the moving of annotations due to gene
merging is still not what I expect.

Before:

entered_by Suparna
entered_date 2004-09-09
updated_by Sue
updated_date 2005-09-05

Expected result:

entered_by Suparna (same)
entered_date 2004-09-09 (same)
updated_by Sue (same as old)
updated_date 2005-09-05 (same as old)

Current result:

entered by Suparna (same)
entered_date 2004-09-09 (same)
updated_by Tanya (new and unexpected)
updated_date 2005-09-28 (new and unexpected)

I do NOT want the updated by and updated dates to change
when the annotation is moved from the old gene model to the
merged gene model.

Please let me know if you need further explanation.

Thanks,

Tanya

[Danny Yoo]

Logged In: YES
user_id=49843

Fix applied; can you test to make sure it's finally working
ok? Sorry about all the back and forth; there are too many
moving interacting pieces. The test url is:

http://tesuque.stanford.edu:9999/pubtest

Thanks!

[Tanya Berardini]

Logged In: YES
user_id=579762

Yahoo! I think you've got it now. Thanks for tackling this
complicated bit, it will really help with tracking.

Tanya

[Danny Yoo]

Logged In: YES
user_id=49843

Closing bug.