Error when running Scaffold guided Binder design.

[Artolomeus]

I get this error When trying to run a scaffolded guided run. When I dont provide something to inference.input_pdb=/inputs/$input_pdb which is not specified in the examples. Seems like a redundancy and it breaks when the installation paths are different (e.g. when running from a docker image in another folder).

Is the default pdb really necessary?

    self.initialize(conf)
  File "/usr/local/lib/python3.9/dist-packages/rfdiffusion/inference/model_runners.py", line 152, in initialize
    self.target_feats = iu.process_target(self.inf_conf.input_pdb, parse_hetatom=True, center=False)
  File "/usr/local/lib/python3.9/dist-packages/rfdiffusion/inference/utils.py", line 615, in process_target
    target_struct = parse_pdb(pdb_path, parse_hetatom=parse_hetatom)
  File "/usr/local/lib/python3.9/dist-packages/rfdiffusion/inference/utils.py", line 521, in parse_pdb
    with open(filename,"r") as f:
FileNotFoundError: [Errno 2] No such file or directory: '/usr/local/lib/python3.9/dist-packages/rfdiffusion/inference/../../examples/input_pdbs/1qys.pdb'

Set the environment variable HYDRA_FULL_ERROR=1 for a complete stack trace.

And when i do give it the same pdb to inference.input_pdb=/inputs/$input_pdb I get

Error executing job with overrides: ['inference.schedule_directory_path=/schedules', 'inference.output_prefix=/outputs/testing_scaffolds_1', 'denoiser.noise_scale_ca=0.5', 'denoiser.noise_scale_frame=0.5', 'scaffoldguided.scaffoldguided=True', 'scaffoldguided.target_pdb=True', 'scaffoldguided.mask_loops=False', 'scaffoldguided.target_path=/inputs/cont_new3_truncated.pdb', 'scaffoldguided.target_ss=/inputs/cont_new3_truncated_ss.pt', 'scaffoldguided.target_adj=/inputs/cont_new3_truncated_adj.pt', 'scaffoldguided.scaffold_dir=/scaffolds/HHH_bc_ss', 'inference.num_designs=10', 'inference.model_directory_path=/models', 'inference.input_pdb=/inputs/cont_new3_truncated.pdb', 'ppi.hotspot_res=[A153,A155,A158,A194]', 'contigmap.contigs=[A60-260/0 30-40]']
Traceback (most recent call last):
  File "/app/RFdiffusion/scripts/run_inference.py", line 84, in main
    x_init, seq_init = sampler.sample_init()
  File "/usr/local/lib/python3.9/dist-packages/rfdiffusion/inference/model_runners.py", line 832, in sample_init
    assert self.target is None, "Giving a target is the wrong way of handling this is you're doing contigs and secondary structure"
AssertionError: Giving a target is the wrong way of handling this is you're doing contigs and secondary structure

Seems like a the handling on the inference.input_pdb=/inputs/$input_pdb needs to be handled in a different way

[nalaterry]

I came across the same issue as well. I was wondering if you were able to resolve it afterward?