- This topic has 6 replies, 2 voices, and was last updated 7 years, 7 months ago by
.
-
-
Hello,
During my work with score function there was a situation when only one omega backbone angle has a change and I noticed some feature. If you set only one omega angle without changing other backbone torsions and then scoring there will be no change in total energy. Moreover, when I drowned myself a bit deeper I have found out that changing two omegas of different residues do the change in energy but still wrongly. Here complete source and output (first for omega term, then for talaris2014) showing the problem, plot also shows output. This feature appears both when you use set_omega function or chaging via DOF_ID. I’m using 2017.08 build.
#include <devel/init.hh>
#include <core/pose/annotated_sequence.hh>
#include <core/pose/util.hh>
#include <core/pose/Pose.hh>
#include <core/scoring/Energies.hh>
#include <core/scoring/ScoreFunction.hh>
#include <core/scoring/ScoreFunctionFactory.hh>
#include <iostream>
int main(int argc, char * argv[])
{
devel::init(argc, argv);
//core::scoring::ScoreFunctionOP scorefx = core::scoring::ScoreFunctionFactory::create_score_function("talaris2014");
core::scoring::ScoreFunctionOP scorefx(new core::scoring::ScoreFunction);
scorefx->set_weight(core::scoring::omega, 1.0);
core::pose::Pose pose1, pose2, pose3, pose4;
std::string peptide_seq = "AAAAA";
core::pose::make_pose_from_sequence(pose1, peptide_seq, *core::chemical::ChemicalManager::get_instance()->residue_type_set(core::chemical::FA_STANDARD));
for (size_t i = 1; i != pose1.total_residue(); i++)
{
pose1.set_phi(i, -135.0);
pose1.set_psi(i, 135.0);
pose1.set_omega(i, 179.
;
}
pose4 = pose3 = pose2 = pose1;
size_t steps = 6;
double range = 180;
double start = 90;
for (size_t i = 0; i != steps + 1; i++)
{
pose1.set_omega(3, start + i * range / steps);
pose1.set_phi(3, -135);
pose2.set_omega(3, start + i * range / steps);
pose2.set_phi(2, -135);
pose3.set_omega(3, start + i * range / steps);
pose3.set_omega(4, start + i * range / steps);
pose4.set_omega(3, start + i * range / steps);
(*scorefx)(pose1);
(*scorefx)(pose2);
(*scorefx)(pose3);
(*scorefx)(pose4);
std::cout << std::scientific << start + i * range / steps << 't'
<< pose1.energies().total_energy() << 't'
<< pose2.energies().total_energy() << 't'
<< pose3.energies().total_energy() << 't'
<< pose4.energies().total_energy() << std::endl;
}
}// core::scoring::omega only
9.000000e+01 8.100120e+01 8.100120e+01 1.620008e+02 8.100120e+01
1.200000e+02 3.600120e+01 8.100120e+01 1.170008e+02 8.100120e+01
1.500000e+02 9.001200e+00 8.100120e+01 9.000080e+01 8.100120e+01
1.800000e+02 1.200000e-03 8.100120e+01 8.100080e+01 8.100120e+01
2.100000e+02 9.001200e+00 8.100120e+01 9.000080e+01 8.100120e+01
2.400000e+02 3.600120e+01 8.100120e+01 1.170008e+02 8.100120e+01
2.700000e+02 8.100120e+01 8.100120e+01 1.620008e+02 8.100120e+01
// talaris2014
9.000000e+01 5.534642e+01 5.534642e+01 1.056326e+02 5.534642e+01
1.200000e+02 2.728259e+01 5.540759e+01 7.771807e+01 5.540759e+01
1.500000e+02 1.054679e+01 5.554679e+01 6.108479e+01 5.554679e+01
1.800000e+02 5.057364e+00 5.568236e+01 5.568270e+01 5.568236e+01
2.100000e+02 1.073223e+01 5.573223e+01 6.137627e+01 5.573223e+01
2.400000e+02 2.754141e+01 5.566641e+01 7.804450e+01 5.566641e+01
2.700000e+02 5.557881e+01 5.557881e+01 1.060839e+02 5.557881e+01
So is that mean that I always need to catch such situations and reset other torsions when dealing with omegas? If I understand scoring and omega term correctly pose 1, 2, and 4 energies must have identical values, or I am not right?
Thanks in advance
-
-
-
-
-
-
-
- You must be logged in to reply to this topic.