How to tell where and what fragment or rotamer was inserted?

Member Site Forums PyRosetta PyRosetta – General How to tell where and what fragment or rotamer was inserted?

Viewing 0 reply threads
  • Author
    Posts
    • #410
      Anonymous

        I’m a new user doing the workshop tutorials.

        first, off there are no .fix_everything() and .set_pack_residue() methods for a packer task. We are guessing that the methods named like those but with the word “temporary” before it are the correct methods. These seem to function.

        our protein is 2ptl (protein L, truncated to 60 residues) from the rosetta benchmarks.

        my first question questions how to query what rotamer(s) or fragments were inserted (and where) after a packer task or a classic fragment move. Which fragment was inserted and where. Seems like there ought to be some sort of “get_last_move()” method lurking or some way to tell which fragment or rotamer was inserted and where.

        the second question is what to expect from a packertask (see page 28 of lulu book where on is inserting at a single rotamer position). when we apply this packer task (pg 28 of the lulu book) and then look at the chi angles they don’t correspond to the chi angles of any dunbrack rotamer. I would have expected it to be a specific rotamer, unless there is also some sort of minimization lurking under the hood.

        print a;
        PDB file name: 2ptl/2ptl.pdb
        Total residues:60
        Sequence: VTIKANLIFANGSTQTAEFKGTFEKATSEAYAYADTLKKDNGEYTVDVADKGYTLNIKFA
        Fold tree:
        FOLD_TREE EDGE 1 60 -1

        packmover.apply(a)
        core.pack.interaction_graph.interaction_graph_factory: Instantiating DensePDInteractionGraph
        core.pack.pack_rotamers: built 41 rotamers at 1 positions.
        core.pack.pack_rotamers: IG: 612 bytes

        let’s look at residue 4, which is a ‘K’

        In [344]: a.chi(4,4)
        Out[344]: -179.54589407726752

        In [345]: a.chi(3,4)
        Out[345]: -68.231809520530959

        In [346]: a.chi(2,4)
        Out[346]: -179.55489877292882

        In [347]: a.chi(1,4)
        Out[347]: -167.63933976788715

        then hunting in the dunbrack lysine file:

        grep 167 dun08/lys.Feb24.Optim.bbdep.regular.lib | grep 179 | grep 68
        LYS -170 -10 1 2 1 2 2 0.016755 -179.3 68.3 175.9 177.5 8.7 10.8 13.0 14.2
        LYS -170 0 1 2 1 2 2 0.016786 -179.3 68.3 175.9 177.5 8.7 10.8 13.0 14.2
        LYS -140 -120 0 2 1 2 2 0.016756 -179.3 68.3 175.9 177.5 8.7 10.9 13.0 14.2
        LYS -60 0 42 3 2 2 2 0.301167 -68.4 179.9 179.7 178.8 6.5 8.7 11.9 12.9
        LYS 40 -90 0 2 1 2 2 0.004167 -179.3 68.3 175.9 177.5 8.7 10.8 13.0 14.2
        LYS 40 130 0 3 2 1 1 0.024167 -67.8 179.0 70.7 68.5 9.5 13.2 15.0 13.4
        LYS 120 -90 0 2 1 2 2 0.016764 -179.3 68.3 175.9 177.5 8.7 10.8 13.0 14.2
        LYS 140 70 0 3 2 1 1 0.011674 -67.8 179.0 70.6 68.5 9.3 13.0 14.9 13.4

        we don’t find any lys rotamers with a -179 chi4 or a -68 chi3 etc…

        this makes one wonder if this is loading up some other database of rotamers or if it’s doing some sort of minimization on the angles in addition to picking a rotamer.

        We notice a possibly related issue on page 27 when we score the position 4 rotamer. the dunbrack energy does not seem to correspond to the log of the dunbrack rotamer probability we are expecting. (it’s close, off by about 0.1)

    Viewing 0 reply threads
    • You must be logged in to reply to this topic.