Rosetta doesn’t read the starting structure – solved

Member Site Forums Rosetta++ Rosetta++ – General Rosetta doesn’t read the starting structure – solved

Viewing 4 reply threads
  • Author
    Posts
    • #320
      Anonymous

        Dear all,

        I’m trying to run Rosetta in relax mode to do refinement of some protein structures that i’ve got from an abinitio run.
        I did the clustering, and i extracted some decoys from the output silent file with Reconstruct_PDB_by_index .
        I used many different lines to do the relaxation but i get an error every time:
        A sample line:

        /home/Bob/Rosetta/rosetta++/rosetta.gcc ac 1py6 a -relax -s 1py6_start_0001.pdb -paths path.txt -nstruct 1


        input_pdb.cc outputs:
        __"cant' find starting residue in pdb file"__

        than it outputs:
        __ERROR:: Unable to determine sequence length from pdb file__

        Than Rosetta exits in :
        __ERROR:: Exit from: pose.cc line: 2013__

        Obviously Rosetta can’t read the starting structure properly. I’ve checked the input pdb files, and they are exactly the same as the files from samples provided with in the RosettaExamples.tgz . I also tried using a true native PDB file.
        Something is wrong with the pdb_stream but i don’t know how to handle it.
        If you have any ideas how to solve the problem, i would be very grateful if you shared it.

        Best Regards
        Bob3k

      • #4041
        Anonymous

          1. -s read the pdb file’s tag but will not understand the extension. you need to remove the extension “.pdb”.
          2. did you specify the paths.txt correctly? is the pdb in the corresponding paths?

          > Dear all,
          >
          > I’m trying to run Rosetta in relax mode to do refinement of some protein structures that i’ve got from an abinitio run.
          > I did the clustering, and i extracted some decoys from the output silent file with Reconstruct_PDB_by_index .
          > I used many different lines to do the relaxation but i get an error every time:
          > A sample line:
          >
          > /home/Bob/Rosetta/rosetta++/rosetta.gcc ac 1py6 a -relax -s 1py6_start_0001.pdb -paths path.txt -nstruct 1
          >
          >
          >
          > input_pdb.cc outputs:
          > __”cant’ find starting residue in pdb file”__
          >
          > than it outputs:
          > __ERROR:: Unable to determine sequence length from pdb file__
          >
          > Than Rosetta exits in :
          > __ERROR:: Exit from: pose.cc line: 2013__
          >
          >
          > Obviously Rosetta can’t read the starting structure properly. I’ve checked the input pdb files, and they are exactly the same as the files from samples provided with in the RosettaExamples.tgz . I also tried using a true native PDB file.
          > Something is wrong with the pdb_stream but i don’t know how to handle it.
          > If you have any ideas how to solve the problem, i would be very grateful if you shared it.
          >
          > Best Regards
          > Bob3k

        • #4042
          Anonymous

            1. I removed the extension but it didn’t make any difference.

            2.I’m pretty much sure that the path.txt file is well specified. I tried to use as a starting structure a file that didn’t exist. Rosetta behaved properly and it output the error that there is no such file.

            The error lines that I get are:

            __reading starting structure: /home/Bob/Rosetta/Protein_pred/1py6/1py6_start/./1py6_start_0001.pdb__

            __can’t find starting residue in pdb file__

            __ERROR:: Unable to determine sequence length from pdb file__

            and later on

            __STOP: You need to fill the fold_tree before calling Pose::set_coords__

            __ERROR:: Exit from: pose.cc line: 2013__

            So Rosetta sees the file but treats it like a wrong format, but I’m 100% sure that the format is fine.

          • #4043
            Anonymous

              Rosetta needs to find the “ATOM” section in the pdb file. You can send your path file, command line and input file to general-support@rosettacommons.org

            • #4044
              Anonymous

                Thanks to the support-team the problem was solved.

                Well, the thing is that all the input files used for generating the models in the first run (fasta, psipred, fragment…) where named: 1py6a.* or in a similar way. After extraction of decoys from the silent file there was no information regarding the chains in the resulting pdbs. So when I tried to relax such structures, Rosetta, despite having read the input files, was unable to process the starting structures (correct me if I’m wrong).

                I changed names of the input files to 1py6_.* and used the following command line

                ROSETTA ac 1py6 _ -relax -nstruct 1 -s 1py6_start -paths path.txt

                Thanks again Yiliu (:biggrin:)

                Best Regards
                Bob3k

            Viewing 4 reply threads
            • You must be logged in to reply to this topic.