IMP.restrainer examples

basic_setup.py

This example shows how to construct a molecular hierarchy consisting of two proteins, and then apply connectivity restraint obtained from pulldown experimental data to these two proteins.

On the following figure, we see an XML representation of the molecular hierarchy.

<!-- File: eg1_representation.xml -->

<Representation>
    <Protein id="Protein1">
        <Chain filename="data/protein1.pdb" selector="CAlpha"/>
        <Chain>
            <Fragment>
                <GeometricShapeRep total_residue="10">
                    <Sphere/>
                </GeometricShapeRep>
            </Fragment>
        </Chain> 
    </Protein>
    <Protein id="Protein2">
        <Chain filename="data/protein2.pdb"/>
    </Protein>
</Representation>

Each level of molecular hierarchy corresponds to an XML tag. Each tag can have optional attributes. In this example, each protein has a unique id. Protein1 has 2 chains. The first chain is constructed from the PDB file using IMP::atom::CAlphaSelector. The second chain is represented by a sphere fragment that is big enough to contain 10 residues. Protein2 is constructed using the default selector, IMP::atom::NonWaterNonHydrogenSelector.

The following figure shows the definition of the restraint.

<!-- File: eg1_restraint.xml -->

<RestraintSet>
    <Pulldown>
        <Restraint>
            <Particle id="Protein1"/>
            <Particle id="Protein2"/>
        </Restraint>
    </Pulldown>
</RestraintSet>

The pulldown restraint is applied to Protein1 and Protein2.

The following Python script demonstrates the process of loading our data into IMP model.

#-- File: basic_setup.py --#

import IMP
import IMP.restrainer

# Load molecular hierarchy definition
rep_parser = IMP.restrainer.XMLRepresentation(
                IMP.restrainer.get_example_path('input/eg1_representation.xml'))
representation = rep_parser.run()

# Load the restraint set
restraint_parser = IMP.restrainer.XMLRestraint(
                     IMP.restrainer.get_example_path('input/eg1_restraint.xml'))
restraint = restraint_parser.run()

# Place representation into IMP model
model = representation.to_model()

# Place restraint into IMP model
restraint.add_to_representation(representation)

###=======================================================================###
#  At this point all data from XML files have been placed into the model.
#  Now it is possible to perform various operations on the IMP model.
###=======================================================================###

model.show()
model.evaluate(False)

rigid_body_and_excluded_volume_restraint.py

This example shows how to use IMP::core::RigidBody and IMP::core::ExcludedVolumeRestraint. We will construct a molecular hierarchy consisting of five proteins, and then apply connectivity restraint obtained from yeast two-hybrid experimental data to these proteins. We will define these proteins as rigid bodies and apply excluded volume restraint to prevent these proteins from interpenetrating. On the following figure, we see an XML representation of the molecular hierarchy.

<!-- File: eg2_representation.xml -->

<Representation>
    <Protein id="Protein1"><Chain filename="data/protein1.pdb"/></Protein>
    <Protein id="Protein2"><Chain filename="data/protein2.pdb"/></Protein>
    <Protein id="Protein3"><Chain filename="data/protein3.pdb"/></Protein>
    <Protein id="Protein4"><Chain filename="data/protein4.pdb"/></Protein>
    <Protein id="Protein5"><Chain filename="data/protein5.pdb"/></Protein>
</Representation>

From restrainer's point of view, a Rigid Body is a special form of restraint. Therefore, to make some parts of the hierarchy rigid, suitable rigid body declarations should be placed in the restraint XML file.

<!-- File: eg2_restraint.xml -->

<RestraintSet>
    <RigidBody>
        <Restraint><Particle id="Protein1"/></Restraint>
        <Restraint><Particle id="Protein2"/></Restraint>
        <Restraint><Particle id="Protein3"/></Restraint>
        <Restraint><Particle id="Protein4"/></Restraint>
        <Restraint><Particle id="Protein5"/></Restraint>
    </RigidBody>
    <ExcludedVolume>
        <Restraint>
            <Particle id="Protein1"/>
            <Particle id="Protein2"/>
            <Particle id="Protein3"/>
            <Particle id="Protein4"/>
            <Particle id="Protein5"/>
        </Restraint>
    </ExcludedVolume>
    <Y2H>
        <Restraint>
            <Particle id="Protein1"/>
            <Particle id="Protein2"/>
            <Particle id="Protein3"/>
        </Restraint>
        <Restraint>
            <Particle id="Protein4"/>
            <Particle id="Protein5"/>
        </Restraint>
    </Y2H>
</RestraintSet>

#-- File: rigid_body_and_excluded_volume_restraint.py --#

import IMP
import IMP.restrainer

# Load molecular hierarchy definition
rep_parser = IMP.restrainer.XMLRepresentation(
                IMP.restrainer.get_example_path('input/eg2_representation.xml'))
representation = rep_parser.run()

# Load the restraint set
restraint_parser = IMP.restrainer.XMLRestraint(
                      IMP.restrainer.get_example_path('input/eg2_restraint.xml'))
restraint = restraint_parser.run()

# Place representation into IMP model
model = representation.to_model()

# Place restraint into IMP model
restraint.add_to_representation(representation)


###=======================================================================###
#  At this point all data from XML files have been placed into the model.
#  Now it is possible to perform various operations on the IMP model.
###=======================================================================###


# Find the IMP data structure with the given id
protein1_hierarchy = representation.get_imp_hierarchy_by_id('Protein1')

# Get root hierarchy
root_hierarchy = representation.get_root_imp_hierarchy()

# Get the rigid body with the given id
protein1_rb = restraint.get_rigid_body_by_id ("Protein1")

# Get all rigid bodies
rbs = restraint.get_all_rigid_bodies()

# Define transformation
ub = IMP.algebra.Vector3D(-50.0,-50.0,-50.0)
lb = IMP.algebra.Vector3D( 50.0, 50.0, 50.0)
bb = IMP.algebra.BoundingBox3D(ub, lb)
translation = IMP.algebra.get_random_vector_in(bb)
rotation = IMP.algebra.get_random_rotation_3d()
transformation = IMP.algebra.Transformation3D(rotation, translation)

# Perform geometric transformations on the Protein1 rigid body
protein1_rb.set_transformation(transformation)
protein1_rb.show()

em_restraint.py

This example shows how to use IMP::em::FitRestraint.

<!-- File: em_representation.xml -->

<Representation>
    <Protein id="Protein1">
        <Chain>
            <Fragment id="frag1">
                <GeometricShapeRep>
                    <Sphere radius="1.0" weight="1.0">
                        <InitialPosition optimize="1" x="12.0" y="12.0" z="12.0"/>
                    </Sphere>
                </GeometricShapeRep>
            </Fragment>
            <Fragment id="frag2">
                <GeometricShapeRep>
                    <Sphere radius="1.0" weight="1.0">
                        <InitialPosition optimize="1" x="15.0" y="6.0" z="6.0"/>
                    </Sphere>
                </GeometricShapeRep>
            </Fragment>
            <Fragment id="frag3">
                <GeometricShapeRep>
                    <Sphere radius="1.0" weight="1.0">
                        <InitialPosition optimize="1" x="6.0" y="15.0" z="15.0"/>
                    </Sphere>
                </GeometricShapeRep>
            </Fragment>
        </Chain>
    </Protein>
</Representation>

<!-- File: em_restraint.xml -->

<RestraintSet>
    <EM>
        <Restraint name="em_restraint" density_filename="data/in.mrc" spacing="1.0" resolution="3.0" xorigin="-6.0" yorigin="-6.0" zorigin="-6.0">
            <Particle id="frag1"/>
            <Particle id="frag2"/>
            <Particle id="frag3"/>
        </Restraint>
    </EM>
</RestraintSet>

#-- File: em_restraint.py --#

import IMP
import IMP.restrainer

# Load molecular hierarchy definition
rep_parser = IMP.restrainer.XMLRepresentation(
                IMP.restrainer.get_example_path('input/em_representation.xml'))
representation = rep_parser.run()

# Load the restraint set
restraint_parser = IMP.restrainer.XMLRestraint(
                      IMP.restrainer.get_example_path('input/em_restraint.xml'))
restraint = restraint_parser.run()

# Place representation into IMP model
model = representation.to_model()

# Place restraint into IMP model
restraint.add_to_representation(representation)


###=======================================================================###
#  At this point all data from XML files have been placed into the model.
#  Now it is possible to perform various operations on the IMP model.
###=======================================================================###


# Get restraint by name
r = restraint.get_restraint_by_name('em_restraint')

# Get EM density map header from the restraint
dmap_header = r.dmap_header

# Get IMP::em::FitRestraint
fit_restraint = r.imp_restraint
fit_restraint.evaluate(False)

saxs_restraint.py

This example shows how to use IMP::saxs::Restraint.

<!-- File: saxs_representation.xml -->

<Representation>
    <Protein id="6lyz">
        <Chain filename="data/6lyz.pdb" selector="NonWaterNonHydrogen"/>
    </Protein>
</Representation>

<!-- File: saxs_restraint.xml -->

<RestraintSet>
    <SAXS> 
        <Restraint name="saxs_restraint" profile_filename="data/lyzexp.dat">
            <Particle id="6lyz"></Particle>
        </Restraint>
    </SAXS>
</RestraintSet>

#-- File: saxs_restraint.py --#

import IMP
import IMP.restrainer

# Load molecular hierarchy definition
rep_parser = IMP.restrainer.XMLRepresentation(
                 IMP.restrainer.get_example_path('input/saxs_representation.xml'))
representation = rep_parser.run()

# Load the restraint set
restraint_parser = IMP.restrainer.XMLRestraint(
                       IMP.restrainer.get_example_path('input/saxs_restraint.xml'))
restraint = restraint_parser.run()

# Place representation into IMP model
model = representation.to_model()

# Place restraint into IMP model
restraint.add_to_representation(representation)


###=======================================================================###
#  At this point all data from XML files have been placed into the model.
#  Now it is possible to perform various operations on the IMP model.
###=======================================================================###


# Get restraint by name
r = restraint.get_restraint_by_name('saxs_restraint')

# Get IMP::saxs::Restraint
saxs_restraint = r.imp_restraint
saxs_restraint.evaluate(False)

basic_display.py

In case we would like to create Chimera logs storing optimization states, we can provide geometric definitions in a display XML file. These XML definitions form a basic interface to IMP::display module. Currently, the only geometric aspect that can be specified is color.

This example shows how to color the representation components and output display file in chimera format.

<!-- File: display_representation.xml -->

<Representation>
    <Protein id="Protein1">
        <Chain id="Chain1" filename="data/6lyz.pdb">
            <Fragment id="Frag1"><AtomicRep start_residue="1" end_residue="10"/></Fragment>
            <Fragment id="Frag2"><AtomicRep start_residue="11" end_residue="20"/></Fragment>
            <Fragment id="Frag3"><AtomicRep start_residue="21" end_residue="30"/></Fragment>
            <Fragment id="Frag4"><AtomicRep start_residue="31" end_residue="40"/></Fragment>
        </Chain>
    </Protein>
</Representation>

<!-- File: pdb_display.xml -->

<Display>
    <Protein id="Protein1">
        <Chain id="Chain1"><Color r="1" g="1" b="1"/>
            <Fragment id="Frag1"><Color r="0" g="1" b="1"/></Fragment>
            <Fragment id="Frag2"><Color r="1" g="0" b="1"/></Fragment>
            <Fragment id="Frag3"><Color r="1" g="1" b="0"/></Fragment>
            <Fragment id="Frag4"><Color r="1" g="0" b="0"/></Fragment>
        </Chain>
    </Protein>
</Display>

Loading a display XML file is similar to loading the representation and restraint XML definitions. Only particles mentioned in the display XML file will be present (this is useful if we only want to display some portion of an hierarchy). The arguments to the create_log are the representation and the file name template (this '%03d' will be replaced by consecutive numbers if we do optimization). This log is intended to be added to an optimization by performing: optimization.add_optimizer_state(log)

#-- File: basic_display.py --#

import IMP
import IMP.restrainer

# Load molecular hierarchy definition
rep_parser = IMP.restrainer.XMLRepresentation(
                IMP.restrainer.get_example_path('input/display_representation.xml'))
representation = rep_parser.run()

# Load color definition
display_parser = IMP.restrainer.XMLDisplay(
                    IMP.restrainer.get_example_path('input/pdb_display.xml'))
display = display_parser.run()

# Place representation into IMP model
model = representation.to_model()

# Write initial display in Chimera format
log = display.create_log(representation, 'pdb_log_%03d.py')
log.write('initial_display_in_chimera.py')

###=======================================================================###
#  At this point all data from XML files have been placed into the model.
#  Now it is possible to perform various operations on the IMP model.
###=======================================================================###

model.show()
model.evaluate(False)