search

steineggerlab / foldseek

Foldseek enables fast and sensitive comparisons of large structure sets.
https://foldseek.com
GNU General Public License v3.0
693 stars 92 forks source link

No 3D alignment shown in webserver viewer #271

Open atn-journal opened 1 month ago

atn-journal commented 1 month ago

Expected Behavior

In web server, the 3D viewer should show the structural alignment between target and query, as well as RMSD and Tm.

Current Behavior

These last few days it only shows the target, no query nor RMSD or Tm.

Steps to Reproduce (for bugs)

Perform search in webserver.

Foldssek Output (for bugs)

n/a

Context

n/a

Your Environment

Windows 10/11, Firefox/Chrome

Screenshot

image

Also, would be nice if in the webserver I could also check RMSD/Tm without having to open the viewer, like it is for sequence identity. Thanks in advance!

milot-mirdita commented 1 month ago

This should only happen in predicted 3Di mode, since there is no query structure in that case.

What are you searching? Could you send me the query, or the URL to this search?

atn-journal commented 1 month ago

Hi. These are the search results. However, I think the issue is related to doing the search by loading an accession number; I tried uploading the PDB myself and it works as usual.

Alexbateman1 commented 1 month ago

I am also seeing this behaviour when I upload .cif files from AF3. Works fine with PDB files.

milot-mirdita commented 1 month ago

Hi. These are the search results. However, I think the issue is related to doing the search by loading an accession number; I tried uploading the PDB myself and it works as usual.

@atn-journal : This result list looks fine to me. The initial screenshots seems to have been from a monomer result, while the link above is a multimer result. Was this intended? Did I miss anything broken?

I am also seeing this behaviour when I upload .cif files from AF3. Works fine with PDB files.

@Alexbateman1: Could you please link me a broken result page? I fixed some AF3 related issues just before release, I might have missed some other though.

milot-mirdita commented 1 month ago

@Alexbateman1: Whoops, I missed something. Same issue as in #274 and likely also the same issue as in the OP.

atn-journal commented 1 month ago

@atn-journal : This result list looks fine to me. The initial screenshots seems to have been from a monomer result, while the link above is a multimer result. Was this intended? Did I miss anything broken?

@milot-mirdita: My bad! Sent the wrong link. Can´t find the other one now. However, I´ve run a new search loading the accession number and now can´t see anything in the viewer, see here. It does work if I upload the PDB, see here.

milot-mirdita commented 1 month ago

I tried deploying a quick fix yesterday night, but realized that the mmCIF support I implemented is still incomplete.

milot-mirdita commented 1 month ago

This should be fully fixed since last Friday. Please let me know if you still have issues with mmCIF input.

Alexbateman1 commented 1 month ago

Yes this issue is now resolved. Thanks for the great work!

On 22 May 2024, at 06:14, Milot Mirdita @.***> wrote:

This should be fully fixed since last Friday. Please let me know if you still have issues with mmCIF input.

— Reply to this email directly, view it on GitHub https://github.com/steineggerlab/foldseek/issues/271#issuecomment-2123881662, or unsubscribe https://github.com/notifications/unsubscribe-auth/AJKGZGOQR4XVCZA6BP3QH2DZDQSTDAVCNFSM6AAAAABHUNAWI6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDCMRTHA4DCNRWGI. You are receiving this because you were mentioned.