PAE Jiffy results not available

[murraydt]

After submitting .crd and .json files, I am not getting any const.inp for download. When I click "submit," I see the wheels turning but shortly afterwards, the "submit" button refreshes and there are no results. - Dan

[dsclassen]

yep... seems to be a bug. Thanks for reporting

[murraydt]

Np. Will await the fix.

On Tue, Oct 3, 2023 at 1:33 PM Scott Classen @.***> wrote:

yep... seems to be a bug. Thanks for reporting

— Reply to this email directly, view it on GitHub https://github.com/bl1231/bilbomd-ui/issues/242#issuecomment-1745682824, or unsubscribe https://github.com/notifications/unsubscribe-auth/A74D7CQ5ECEUF6YJV2ARW5DX5RZC7AVCNFSM6AAAAAA5RQZULSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTONBVGY4DEOBSGQ . You are receiving this because you authored the thread.Message ID: @.***>

[dsclassen]

Where did the PAE JSON file come from?

[dsclassen]

For posterity:

For a protein of length num_res, the JSON file has the following structure of arrays format:

[
    {
        "predicted_aligned_error": [[0, 1, 4, 7, 9, ...], ...], # Shape: (num_res, num_res).
        "max_predicted_aligned_error": 31.75
    }
]

The fields in the JSON file are:

• predicted_aligned_error: The PAE value of the residue pair, rounded to the closest integer. For the PAE value at position (i, j), i is the residue on which the structure is aligned, j is the residue on which the error is predicted.

• max_predicted_aligned_error: A number that denotes the maximum possible value of PAE. The smallest possible value of PAE is 0.

PAE file formats are not defined so I suppose the output from other (or older) versions of Alphafold might cause problems.

[dsclassen]

I think the problem PAE JSON file has this organization:

{
    "max_pae": 31.75, "pae": [[0.75, 1.88, 4.42,...], ...], # Shape: (num_res, num_res).
    "ptm": 0.91,
    "iptm": 0.9100000262260437
}

[murraydt]

That JSON file came from an earlier folding run I did back when the AF2 machine was just set up. I will try another run and see what happens.

On Tue, Oct 3, 2023 at 1:50 PM Scott Classen @.***> wrote:

I think the problem PAE JSON file has this organization:

{ "max_pae": 31.75, "pae": [[0.75, 1.88, 4.42,...], ...], # Shape: (num_res, num_res). "ptm": 0.91, "iptm": 0.9100000262260437 }

— Reply to this email directly, view it on GitHub https://github.com/bl1231/bilbomd-ui/issues/242#issuecomment-1745704771, or unsubscribe https://github.com/notifications/unsubscribe-auth/A74D7CVUFQGAADOGIYCR2I3X5R3AJAVCNFSM6AAAAAA5RQZULSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTONBVG4YDINZXGE . You are receiving this because you authored the thread.Message ID: @.***>

[dsclassen]

I know that at some point Alphafold changed the format of the output PAE files, but your uploaded file doesn't really match what I think of as a "legacy" PAE JSON file.... weird. Regardless, I should implement a file format check and not just leave the user wondering why shit isn't working.

[murraydt]

I'm excited to use this feature to generate const.inp files. This particular complex doesn't have a published structure and my early folding prediction agreed well with the SAXS, so I am interested to see what moves as it goes between states. It has a ton of clusters and an FMN (pic), but AF2 doesn't show their coordinations. I won't be able to use BilboMD to see cofactor proximities, but I will at least see any movements. My folding run is running now.

On Tue, Oct 3, 2023 at 1:57 PM Scott Classen @.***> wrote:

I know that at some point Alphafold changed the format of the output PAE files, but your uploaded file doesn't really match what I think of as a "legacy" PAE JSON file.... weird. Regardless, I should implement a file format check and not just leave the user wondering why shit isn't working.

— Reply to this email directly, view it on GitHub https://github.com/bl1231/bilbomd-ui/issues/242#issuecomment-1745714013, or unsubscribe https://github.com/notifications/unsubscribe-auth/A74D7CQOOLZDRG7VC5YAT2TX5R32VAVCNFSM6AAAAAA5RQZULSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTONBVG4YTIMBRGM . You are receiving this because you authored the thread.Message ID: @.***>

[dsclassen]

@michalhammel Can you comment on this? Do you think your current pae_ratios.py script can handle this old PAE format than Dan tried to upload to the PAE Jiffy?

[dsclassen]

Closing for now. We may need to revisit this at some point when we better understand all the different places that PAE results come from and the various file formats.