Score2 and Score5…aren’t hey meant to be different?

Member Site Forums PyRosetta PyRosetta – General Score2 and Score5…aren’t hey meant to be different?

Viewing 1 reply thread
  • Author
    Posts
    • #1520
      Anonymous

        Hello everybody

        doing
        score_fxn5=create_score_function( ‘score5’ )
        score_fxn2=create_score_function( ‘score2’ )

        and printing them I obtain the same output, I thought they were meant to be different.

        print score_fxn?
        ScoreFunction::show():
        weights: (vdw 1) (cenpack 0.5) (pair 1) (env 1) (cbeta 0.25) (hs_pair 1) (ss_pair 1) (sheet 1)
        energy_method_options: EnergyMethodOptions::show: etable_type: FA_STANDARD_DEFAULT
        EnergyMethodOptions::show: unfolded_energies_type: UNFOLDED_SCORE12
        EnergyMethodOptions::show: atom_vdw_atom_type_set_name: centroid
        EnergyMethodOptions::show: exclude_protein_protein_hack_elec: false
        EnergyMethodOptions::show: exclude_monomer_hack_elec: false
        EnergyMethodOptions::show: hackelec_max_dis: 5.5
        EnergyMethodOptions::show: hackelec_min_dis: 1.5
        EnergyMethodOptions::show: hackelec_die: 10
        EnergyMethodOptions::show: hackelec_no_dis_dep_die: false
        EnergyMethodOptions::show: exclude_DNA_DNA: true
        EnergyMethodOptions::show: cst_max_seq_sep: 18446744073709551615
        EnergyMethodOptions::show: bond_angle_central_atoms_to_score:
        EnergyMethodOptions::show: bond_angle_residue_type_param_set: none
        HBondOptions::show: exclude_DNA_DNA: true
        HBondOptions::show: include_intra_res_RNA_: false
        HBondOptions::show: exclude_self_hbonds: true
        HBondOptions::show: use_hb_env_dep: true
        HBondOptions::show: use_hb_env_dep_DNA: true
        HBondOptions::show: smooth_hb_env_dep: true
        HBondOptions::show: bb_donor_acceptor_check: true
        HBondOptions::show: decompose_bb_hb_into_pair_energies: false
        HBondOptions::show: params_database_tag_: standard_params
        HBondOptions::show: use_incorrect_deriv_: false
        HBondOptions::show: use_sp2_chi_penalty_: false
        HBondOptions::show: Mbhbond: false

        I thought they were meant to be different.

      • #8435
        Anonymous

          Here are the weights files from the scoring database:

          rosetta_database/scoring/weights> cat score2.wts
          # score2 from rosetta++, used in stage 3 of ClassicAbinitio protocol.
          env 1.0
          pair 1.0
          cbeta 0.25
          cenpack 0.5
          vdw 1.0
          rg 0.0
          hs_pair 1.0
          ss_pair 1.0
          rsigma 0.0
          sheet 1.0

          STRAND_STRAND_WEIGHTS 1 6

          rosetta_database/scoring/weights> cat score5.wts
          # score5.wts, used in stage 3 of ClassicAbinitio protocol
          env 1.0
          pair 1.0
          cbeta 0.25
          cenpack 0.5
          hs_pair 1.0
          ss_pair 1.0
          rsigma 0.0
          sheet 1.0
          rg 0.0
          vdw 1.0

          STRAND_STRAND_WEIGHTS 1 11

          STRAND_STRAND_WEIGHTS is different, whatever that is. I’ll see if I can dig out what that does.

        • #8436
          Anonymous

            Apparently scores 2 and 5 are both for stage 3 of ClassicAbinitio (as it says). Here’s the code comment:

            – Stage 3: uses large randomly selected fragment insertions, although the size of the fragment insertions
            is tunable via the set_apply_large_frags( bool ) method. Alternates between score2.wts_patch and score5.wts_patch,
            running tunable numbers of 2000-cycle iterations between the two scoring functions.

            The STRAND_STRAND_WEIGHTS line ultimately weasels its way to the class SecondaryStructurePotential, which controls hs_pair, ss_pair, rsigma, and sheet. I *think* those two numbers are the distance, in angstroms, bounding how far apart two strands can be and still be considered in a “sheet”, but I’m not at all confident of that. This chunk of code was originally written in FORTRAN more than a decade ago, so it’s gotten a little obscure over the years. Look at SecondaryStructurePotential::sspair to see them in action (as lowstrand and cutoff).

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