[lgd_cluster_bsas.py] Clustering has failed.

[lightdock]

[lgd_cluster_bsas.py] Clustering has failed.

[SBalza]

Unfortunately, I can't share these structures specifically. I will check them individually and try to figure out why they might have different atoms.
The tutorial structures are working fine which suggests mine have some anomaly. While I review them, I am going to run them through PDB Fixer and re-try.

[SBalza]

As you say that I was reading the script and realized. Me messing with the .sh script had edited a few pdb. That is exactly what fixed it! Thank you for the help.

[brianjimenez]

Hi @SBalza . I see different input structures here:

lgd_generate_conformations.py ../scFv014.pdb ../CD11b_aADomain.pdb

and here:

lightdock3_setup.py --noh --noxt rec.pdb lig.pdb -r restraints_file

Can you share setup.json, lightdock.info and input PDB structures if possible? Have you tried the tutorial structures?

[brianjimenez]

From the output provided, several structures are clustered ightdock_68.pdb, ightdock_193.pdb, etc., but ightdock_1.pdb seems to be corrupted somehow. Did you try to open that PDB file in a text editor/molecular visualizer to see if the content is OK?

[SBalza]

Hi @brianjimenez,
Thanks for the fast response. I renamed it as I was typing it out but yes those are the original structures, they match throughout the execution. setup.json and lightdock.info below.

I had not opened it, but checking now lightdock1.pdb on Pymol, all chains present and looks fine:

setup.json:

{
    "anm_lig": 10,
    "anm_lig_rmsd": 0.5,
    "anm_rec": 10,
    "anm_rec_rmsd": 0.5,
    "anm_seed": 324324,
    "dense_sampling": false,
    "fixed_distance": 0.0,
    "flip": false,
    "glowworms": 200,
    "ligand_pdb": "lig.pdb",
    "ligand_restraints": {
        "active": [],
        "blocked": [],
        "passive": []
    },
    "membrane": false,
    "noh": true,
    "now": false,
    "noxt": true,
    "receptor_pdb": "rec.pdb",
    "receptor_restraints": {
        "active": [
            "H.SER.97",
            "H.SER.98",
            "H.GLY.99",
            "H.VAL.100",
            "H.SER.30"
        ],
        "blocked": [],
        "passive": []
    },
    "restraints": "restraints_file",
    "setup_version": "0.9.4",
    "starting_points_seed": 324324,
    "surface_density": 50.0,
    "swarm_radius": 10.0,
    "swarms": 100,
    "swarms_per_restraint": 20,
    "transmembrane": false,
    "use_anm": false,
    "verbose_parser": false,
    "write_starting_positions": false

lightdock.info:

{
    "anm_lig": 10,
    "anm_lig_rmsd": 0.5,
    "anm_rec": 10,
    "anm_rec_rmsd": 0.5,
    "anm_seed": 324324,
    "configuration_file": null,
    "cores": null,
    "dense_sampling": false,
    "fixed_distance": 0.0,
    "flip": false,
    "glowworms": 200,
    "gso_seed": 324324,
    "ligand_pdb": "lig.pdb",
    "ligand_restraints": {
        "active": [],
        "blocked": [],
        "passive": []
    },
    "listscoring": null,
    "local_minimization": false,
    "membrane": false,
    "mpi": false,
    "nmodes_step": 0.5,
    "noh": true,
    "now": false,
    "noxt": true,
    "profiling": false,
    "receptor_pdb": "rec.pdb",
    "receptor_restraints": {
        "active": [
            "H.SER.97",
            "H.SER.98",
            "H.GLY.99",
            "H.VAL.100",
            "H.SER.30"
        ],
        "blocked": [],
        "passive": []
    },
    "restraints": "restraints_file",
    "rotation_step": 0.5,
    "scoring_function": null,
    "setup_file": "setup.json",
    "setup_version": "0.9.4",
    "simulation_version": "0.9.4",
    "start_time": "2024-01-26 13:07:03",
    "starting_points_seed": 324324,
    "steps": 10,
    "surface_density": 50.0,
    "swarm_list": null,
    "swarm_radius": 10.0,
    "swarms": 100,
    "swarms_per_restraint": 20,
    "translation_step": 0.5,
    "transmembrane": false,
    "use_anm": false,
    "verbose_parser": false,
    "write_starting_positions": false

[SBalza]

Would it be worthwhile to edit lgd_cluster_bsas.py to catch/handle the error to write the ones that do cluster anyways?
I am not sure how one structure could be corrupted but the others seem to be clustering.

[brianjimenez]

It is kind of weird indeed. If you could share the input structures I'd try to reproduce it locally. The error to me looks like there is a different number of CA atoms between the two structures to be clustered, thus raising that exception.

[brianjimenez]

No problem! Try removing and cleaning all simulation and analysis data generated by LightDock and try again from the very beginning. Maybe if you run multiple times with different parameters without cleaning up swarm folders, lgd_generate_conformations.py script appends PDB data to the same PDB file (this is the way to add ligand PDB structure after receptor PDB data is written to lightdock_XXX.pdb).

[brianjimenez]

No problem, all's well that ends well!