What does it mean when there is no output data in 'final_design_stats.csv'?

[nduan1]

Hi, the pipeline is amazing! I am trying to run a prediction on a ~300-aa peptide. Everything seems to be running smoothly, but there is no output in the 'final_design_stats.csv' file or the 'Accepted' folder after 6 days of processing. It is still running, though.

There is some data in the 'mpnn_design_stats.csv' and 'trajectory_stats.csv' files.

I used the default settings on my command line.
Here is the content of my JSON file:

{
"design_path": "/data/EmiolaLab/duann2/bin_proj/bindcraft/bcraft/BindCraft/candidate4-2/",
"binder_name": "can4",
"starting_pdb": "/data/EmiolaLab/duann2/bin_proj/bindcraft/bcraft/BindCraft/candidate4-2/can4.pdb",
"chains": "A",
"target_hotspot_residues": null,
"lengths": [1, 289],
"number_of_final_designs": 100
}

What does it mean if there is no final output in the 'final_design_stats.csv' file or the 'Accepted' folder? Should I change any settings? I would greatly appreciate any instructions you can provide.

Thanks!

[LennartNickel]

This means that no designs have made it through the filters. Have a look at the trajectories being sampled and see if they are reasonable or if you can find a reason for the failure in the failure.csv file. This can help you to adjust the settings. Sometimes, a lot of sampling is necessary, and in rare cases it just doesn't find a solution. However, I would advice to try a few more inputs (af2 prediction, difference crystal structures) and different hotspots.