Open nataled opened 3 years ago
From Julie: I can confirm that when loading the very slow msa for https://proconsortium.org/app_test/entry/PR:000036285/ that has 230 rows there is only one sequence for P49768 so muscle doesn't actually run. It does take some time for the server to respond with the annotated data it returns (which includes all the details of the ptms and sequences) so it is probably worth optimizing that code too in addition to the angularjs code that the browser runs over that data that actually generates the html.
This page times out: https://proconsortium.org/app_test/entry/PR:P20039
Unclear if that's related to decoration since there is no UniProtKB accession indicated on the definition line.
Just temporary. It seems to happen randomly that nothing loads for a few minutes and then goes back to normal. Works now but it returns Sorry, the server is not able to return the alignment because there is no PRO-proteoform-std details in the definition line.
On Fri, Apr 9, 2021 at 2:52 PM Darren A. Natale @.***> wrote:
This page times out: https://proconsortium.org/app_test/entry/PR:P20039
Unclear if that's related to decoration since there is no UniProtKB accession indicated on the definition line.
— You are receiving this because you were assigned. Reply to this email directly, view it on GitHub https://github.com/PROconsortium/PRoteinOntology/issues/232#issuecomment-816890016, or unsubscribe https://github.com/notifications/unsubscribe-auth/AGAYKJXK5FDF3WWS5HBLQATTH5EH3ANCNFSM42VOIRNQ .
All the decoration is done within the client browser--this is slow. The database retrieval is also slow. Will need to look into how these each can be optimized.
Page loading for terms with a lot of sequences to align is very slow. https://proconsortium.org/app/entry/PR%3A000036285/ is a good example. All the sequence variants use the same reference sequence, so no alignment is necessary. It must be the decoration that is slow. This needs to be optimized.