AtomTree::torsion_angle() can’t find dof! in match and enzdes

Member Site Forums Rosetta 3 Rosetta 3 – General AtomTree::torsion_angle() can’t find dof! in match and enzdes

Viewing 1 reply thread
  • Author
    Posts
    • #1437
      Anonymous

        I am running matcher and enzdes using a complex ligand. I keep getting the following lines repeating throughout my output files in both matching and enzdes:

        core.kinematics.AtomTree: AtomTree::torsion_angle() cant find dof! atomno= 6 rsd= 5 atomno= 9 rsd= 5 atomno= 10 rsd= 5 atomno= 11 rsd= 5
        core.kinematics.AtomTree: AtomTree::torsion_angle() cant find dof! atomno= 6 rsd= 5 atomno= 9 rsd= 5 atomno= 14 rsd= 5 atomno= 15 rsd= 5
        core.kinematics.AtomTree: AtomTree::torsion_angle() cant find dof! atomno= 6 rsd= 5 atomno= 9 rsd= 5 atomno= 10 rsd= 5 atomno= 11 rsd= 5
        core.kinematics.AtomTree: AtomTree::torsion_angle() cant find dof! atomno= 6 rsd= 5 atomno= 9 rsd= 5 atomno= 14 rsd= 5 atomno= 15 rsd= 5

        Match seems to produce reasonable results despite the errors. What do they mean? Should I be concerned and try to fix them? Where should I start looking to diagnose the cause of the problem?

        Thanks!

      • #7965
        Anonymous

          The proximal source of the error is in the AtomTree::torsion_angle() function in rosetta_source/src/core/kinematics/AtomTree.cc. I’m not sure if that’ll help you track down the issue though. The key thing you’ll want to figure out is what torsion angle it’s referring to. What’s the residue type for residue 5, and can you figure out what atoms 6, 5, 10, 11, 14 and 15 are?

          If it’s a problem or not depends on where this value is being used. If the torsion isn’t found, the function returns zero, which may or may not cause problems for the calling function, depending on what it’s expecting. As a developer, I would likely track this down by compiling Rosetta in debug mode, running the program that’s giving the issue in a debugger, setting a breakpoint at the error message, and then looking at the backtrace when the program pauses, seeing where the torsion_angle() is being called. – But if you’re getting results and they look to be reasonable, that may be more effort than you’re willing to put in.

      Viewing 1 reply thread
      • You must be logged in to reply to this topic.