Final output is showing 0 for failed and 0 for successful designs

[bifxcore]

I got the whole boltzdesign steps to complete with no error, but the output log is odd:

/ligandmpnn_cutoff_4/03_af_pdb_success/rosetta_energy.csv
Number of designs passing all filters: 0
Number of failed designs: 0
success_sample_num 0
Number of Success designs 0
Rosetta energy calculation completed!
Pipeline completed successfully!

None of the designs passed, but none of them failed... that can't be right?

According to the mask definition in pyrosetta_utils.py#L459 all my designs failed due to larger surface hydrophobicity than the 0.35 threshold.

Number of failed designs should have been: 10.

Code is bugged:

    filtered_df = df[mask].copy()
    failed_df = [] # this is hard coded as empty?
    
    print(f"Number of designs passing all filters: {len(filtered_df)}")
    print(f"Number of failed designs: {len(failed_df)}") # always returns 0

[yehlincho]

Oh, got it. I’ll fix it soon. Potentially, one way might be to redesign the interface region using ProteinMPNN if hydrophobicity is an issue.