Skip to content

Latest commit

 

History

History
540 lines (431 loc) · 22.2 KB

doc_genpoly_lt.md

File metadata and controls

540 lines (431 loc) · 22.2 KB

genpoly_lt.py

DNA model space-filling curve DNA wrapped along curve

Description

genpoly_lt.py is a program for preparing simulations containing polymers. It is useful for wrapping polymers around an arbitrary curve. (An example of its use is provided here.) genpoly_lt.py generates a moltemplate file (.LT file) containing a definition of a polymer molecule containing monomers located at the positions specified in a 3-column text file ("coords.raw" in the example below). Monomers will be rotated so that they point along the polymer axis direction (see "-dir-indices") with an optional helical twist added (see "-helix"). Users can specify one or more bonds connecting each monomer to the next monomer (see "-bond"). Similarly, 3-body and 4-body angular interactions between atoms in different monomers can either be generated automatically (using the standard moltemplate "Angle By Type" rules) OR generated manually (using "-angle", "-dihedral", "-improper" arguments).

Note: This program is both a stand-alone executable program (that can be run from the terminal) and a python module. The former is documented below. (The python API is explained later.)

Usage:

   genpoly_lt.py  \
      [-polymer-name pname] \
      [-monomer-name mname] \
      [-sequence sequence.txt] \
      [-bond a1 a2] \
      [-bond btype a1 a2] \
      [-angle    atype a1 a2 a3 i1 i2 i3] \
      [-dihedral dtype a1 a2 a3 a4 i1 i2 i3 i4] \
      [-improper itype a1 a2 a3 a4 i1 i2 i3 i4] \
      [-inherits ForceFieldObject] \
      [-header "import monomer.lt"] \
      [-helix deltaphi] \
      [-helix-angles helix_angles_file.txt] \
      [-orientations orientations_file.txt] \
      [-quaternions quaternions_file.txt] \
      [-axis x,y,z] \
      [-circular yes/no/connected] \
      [-cuts cuts.txt] \
      [-polymer-directions polarities.txt] \
      [-dir-indices ia ib] \
      [-padding paddingX,paddingY,paddingZ] \
      [-in coords.raw] \
      < coords.raw > polymer.lt

Arguments [optional]

-polymer-name name

Name of the moltemplate object that will be created. (By default "Polymer")

-monomer-name name

Name of the moltemplate object that will be replicated along the length of the polymer(s). ("Monomer" by default). This monomer should be defined elsewhere and ORIENTED SO THAT THE POLYMER AXIS LIES IN THE +X DIRECTION. You can use the "-header" argument to specify where the monomer(s) is defined. *Note: If you are defining heteropolymers or polymers with end-caps, then do not use the "-monomer" argument. Use the "-sequence" argument instead.

You can include rotations or transformations to the monomer subunit before it is moved into position. For example, it is often useful to to use a modified version of the monomer whose initial coordinates are compressed to avoid collisions with other monomers. To do this, use something like "Monomer.scale(0.5,0.7,0.7)" instead of "Monomer". This would compress each monomer lengthwise by 0.5 and 0.7 laterally. (After minimization, each monomer should expand back to its ordinary size and shape.)*

-sequence sequence.txt

If you are building a heteropolymer, this argument allows you to specify the sequence of monomers in the polymer. You can also use this argument to add end-caps (ie custom monomer types) to the ends of your polymer, and orient them in the forward and backward directions. See example below. The "sequence.txt" file contains the sequence of monomers you want in your polymer(s). Each line of this file should be the name of a moltemplate object for the monomer subunit you want at that location. The number of lines in this file should match the sum of all of the lengths of the polymers (which equals the number of lines in the coordinate file). Each type of monomer listed must be a moltemplate object which contains atoms whose $atom (atom-ID) variables match the a1,a2 atoms mentioned in the -bond, -angle, -dihedral, and -improper arguments (if applicable). (In the butane example below, it would be the carbon atom in the backbone.) As before, you can include coordinate transforms in each monomer's name. Here is an example "sequences.txt" file for butane:

  CH3
  CH2
  CH2
  CH3.rot(180,0,0,1)

The "CH2" and "CH3" moltemplate objects are presumably defined elsewhere and ORIENTED WITH THE POLYMER AXIS ALONG THE +X DIRECTION. The ".rot(180,0,0,1)" makes sure the final CH3 monomer is oriented in the -X (opposite) direction. (Additional movement and rotation commands will be added to align each monomer with the direction of the curve.) If you are using the "-cuts" argument to create multiple polymers, then this file would resemble the file above, with the sequence of multiple such polymers appended together. It would include additional "CH3" and "CH3.rot(180,0,0,1) end-cap monomers at places which are before and after the integers specified using the "-cuts" argument.

-bond btype a1 a2

Add a bond between successive monomers of type btype. between atoms named a1 and a2 (all three arguments are strings and omit the @bond: and $atom: prefixes in moltemplate variables) (Note: The bond's type, btype, can be omitted. See below.) Multiple bonds between successive monomers can be added by having "-bond btype a1 a2" appear several times in the argument list. For example, double-stranded DNA can be implemented as a polymer with 2 bonds connecting separate monomers (if each "monomer corresponds to a base pair). If you want to add bonds between atoms in non-consecutive monomers, then you can instead use the genpoly_modify_lt.py program to add modifications to the polymer later. (That program supports the "-bond btype a1 a2 i1 i2" argument allowing you to specify monomer indices i1, i2.)

-bond a1 a2

Add a bond between successive monomers between atoms named a1, a2. Here we omit the bond's type. Omitting the bond's type is allowed if you are using a force field (like OPLSAA), which looks up bond types according to rules defined in the force field.

-angle atype a1 a2 a3 i1 i2 i3

Add a 3-body angle interaction between atoms a1 a2 a3 in monomers i1 i2 and i3. (The atype a1, a2, a3 arguments are strings containing moltemplate variable names. The standard moltemplate prefixes "$angle:", "@angle:", and "$atom:" should be omitted. The i1, i2, i3 arguments are integer indices indicating the monomer that each atom belongs to.

  • 0 corresponds to the current monomer
  • 1 corresponds to the next monomer
  • 2 corresponds to the following monomer, etc...

(For circular polymers, the indices will be wrapped appropriately.) Multiple angles per monomer can be added by having: "-angle aname atype a1 a2 a3 i1 i2 i3" appear several times in the argument list with different parameters (NOTE: USUALLY THE "-angle" ARGUMENT IS NOT NEEDED IF YOU ARE USING A FORCE FIELD LIKE OPLSAA WHICH GENERATES ANGLE INTERACTIONS.)

-dihedral dtype a1 a2 a3 a4 i1 i2 i3 i4

Add a 4-body dihedral interaction between atoms a1 a2 a3 a4 in monomers i1 i2 and i3. (The dtype a1, a2, a3, a4, arguments are strings containing moltemplate variable names. The moltemplate prefixes "$dihedral:", "@dihedral:", and "$atom:" should be omitted The i1, i2, i3, i4 arguments are integer indices indicating the monomer that each atom belongs to. (See explanation above.) Multiple dihedrals per monomer can be added by having: "-dihedral dname dtype a1 a2 a3 a4 i1 i2 i3 i4" appear several times in the argument list with different parameters (NOTE: USUALLY THE "-dihedral" ARGUMENT IS NOT NEEDED IF YOU ARE USING A FORCE FIELD LIKE OPLSAA WHICH GENERATES DIHEDRAL INTERACTIONS.)

-improper itype a1 a2 a3 a4 i1 i2 i3 i4

Add a 4-body improper interaction between atoms a1 a2 a3 a4 in monomers i1 i2 and i3. (The itype a1, a2, a3, a4, arguments are strings containing moltemplate variable names. The moltemplate prefixes "$improper:", "@improper:", and "$atom:" should be omitted The i1, i2, i3, i4 arguments are integer indices indicating the that each atom belongs to. (See explanation above.) Multiple impropers per monomer can be added by having: "-improper iname itype a1 a2 a3 a4 i1 i2 i3 i4" appear several times in the argument list with different parameters (NOTE: USUALLY THE "-improper" ARGUMENT IS NOT NEEDED IF YOU ARE USING A FORCE FIELD LIKE OPLSAA WHICH GENERATES IMPROPER INTERACTIONS, ...OR IF YOUR POLYMER DOES NOT CONTAIN BACKBONE IMPROPERS.)

-inherits FORCE_FIELD

This allows you to add "inherits FORCE_FIELD" when defining the polymer object(s). (It simply adds the text "inherits FORCE_FIELD" after the polymer name.) In this example, "FORCE_FIELD" is the name of a moltemplate object which defines any rules for creating angles, dihedrals, impropers which you want to be generated automatically. This FORCE_FIELD object can be defined elsewhere (such as in a separate file and imported using the "-header" argument).

-header 'some text'

This is a convenient way to insert a line of text at the beginning of the file that will be created by genpoly_lt.py. You can put any text at the beginning of the file, but typically these are "import" statements. For example:

  -header 'import "force_field.lt" # (<--defines FORCE_FIELD)'

or

  -header 'import "monomer.lt"     # (<--defines Monomer)'

The imported .LT files typically contain definitions of monomers or force field parameters that the polymer needs. (They must appear at the beginning of the .LT file, before the polymer is defined or moltemplate will complain later.) As shown in the example, you can insert multiple lines of text at the beginning by using multiple -header arguments.

-cuts cut_locations.txt

Cut the polymer in several places along its length. This is useful if your goal is to create many polymers of different lengths instead of one long polymer. This will simply cut the polymer N times along its length. The file "cut_locations.txt" is a text file containing a list of positive integers (one per line) indicating where you would like the polymer to be cut. For each integer, i, which appears in this file, a cut is made between monomers i-1 and i (Indexing begins at 0, so a value of 1 corresponds to a cut between the first and second monomers.) A separate molecule object will be created for each polymer, and an integer suffix will be added to each name, to distinguish them from each other. (Each of these polymers will be part of a larger object defined by this program. Instantiating that object will create all of the individual polymers.)

NOTE To put end-caps at the ends of each polymer (ie. to change the monomer type at the ends of each polymer), you must use the "-sequence" argument. You must supply a text file with the monomers you want to put at the beginning and ending of each polymer listed at the appropriate place in this file. (You also have the option to apply different rotations to the monomers at either end of each polymer to orient them in the forward and backward directions.) See the description of the -sequence argument for details.

-axis x,y,z direction of the polymer axis in the original monomer object.

These three numbers (separated by commas with no spaces) define the direction that the monomer subunit is pointing in.
By default, the three numbers are 1 0 0 (ie, the X axis)

-helix deltaphi

Optionally, rotate each monomer around it's axis by angle deltaphi (in degrees) beforehand

-helix-angles helix_angles_file.txt

Optionally, rotate each monomer around it's axis by specifying a list of angles contained in a file (eg "helix_angles_file.txt"). This file contains one number per line (one line per monomer). Each number represents the angle of that monomer relative to the previous monomer around that axis.

-circular keyword

Specify whether the polymer is circular. If so, then you must use "-circular yes" or "-circle connected". "keyword" must be one of these choices:

"no" The polymer is a linear chain with the two ends
not connected. (default)
"yes" The polymer is a circular loop with the two ends
connected (and the direction of the last monomer
points toward the first monomer).
"connected" Connect the two ends together with bonds (and angles,
and dihedrals, if applicable) to make a closed loop.
But do not adjust the orientation of the first and
last monomers so that they point towards eachother.
(Use this if you plan to simulate an "infinitely"
long polymer using periodic boundary conditions,
with the two ends are connected on opposite sides.)

-padding paddingX,paddingY,paddingZ

This will cause the program to attempt to estimate the size of the smallest rectangular box which encloses all of the coordinates in the coordinate file. The user must supply 3 comma-separated numbers (no spaces) which indicate how much extra room is needed in the ±x, ±y, ±z directions.

-polymer-directions polarities.txt

Change the order that coordinates are read from the file. This is specified once per polymer. You must supply a file containing one line per polymer. (Unless you used the -cuts argument this file will have only line.) Each line must contain either "1" or "-1". A value of "1" indicates that you want to read the coordinates for that polymer in the order they appear in the coordinate file. (IE. the normal behavior.) A value of -1 will cause the coordinates for that polymer to be reversed after reading. (In other words, read the coordinates from the corresponding portion of the file in reverse order. This feature is probably not useful to most users.)

-dir-indices ia ib

The program attempts to orient each monomer in a direction that the polymer is pointing. By default, the program will orient monomer i in the direction of a line connecting the monomers before and after it (monomers i-1 and i+1). The user can override this using the -dir-indices command line argument. The ia and ib arguments are integer offsets. To point monomer i in the direction connecting it to the following monomer (i+1), use -dir-indices 0 1. arguments are integer offsets. To point monomer i in the direction connecting it to the previous monomer (i-1), use -dir-indices -1 0. (Note: If the -polymer-directions argument is used, and the current polymer has a direction of -1, the indices ia, ib will be flipped.) For circular polymers, the indices will be wrapped appropriately.

-orientations orientations_file.txt

Specify the orientation of each monomer in the polymer by providing a 9-column text file containing a list of rotation matrices (R, one for each monomer). Each 3x3 matrix describes the orientation of that monomer relative to that monomer's original orientation (not relative to to the previous monomer's orientation). Each line in the file contains 9 numbers which are the entries of a matrix:

    R_11, R_12, R_13, R_21, R_22, R_23, R_31, R_32, R_33

This matrix applies the following coordinate transformation:

    /x'\   / R_11 R_12 R_13 \ /x\
    |y'| = | R_21 R_22 R_23 | |y|
    \z'/   \ R_31 R_32 R_33 / \z/

(This can be any linear transformation, not just a rotation.) Note: These transformations are applied after:

  1. ...coordinate transformations included in the monomer's name specified in the "-monomer-name" or "-sequence" arguments, for example: "EthyleneGlycol.move(0.2,-0.7,0).rot(180,1,0,0)"

  2. ...rotations around the axis specified by the "-helix" or "-helix-angles" and "-axis" arguments (if applicable). (Consequently this argument can be supplied together with the "-helix", "-helix-angles", and "-axis" arguments.)

-quaternions quaternions_file.txt

Specify the orientation of each monomer in the polymer by providing a 4-column text file containing a list of quaternions (one for each monomer). Each quaternion describes the orientation of that monomer relative to that monomer's original orientation (not relative to to the previous monomer's orientation). Each quaternion has 4 numbers. The first number is cos(θ/2) (where θ is the rotation angle). The remaining 3 numbers form a vector (of length sin(θ/2)), pointing along the axis of rotation. Note: These rotations are applied after:

  1. ...coordinate transformations included in the monomer's name specified in the "-monomer-name" or "-sequence" arguments, for example: "EthyleneGlycol.move(0.2,-0.7,0).rot(180,1,0,0)"

  2. ...rotations around the axis specified by the "-helix" or "-helix-angles" and "-axis" arguments (if applicable). (Consequently this argument can be supplied together with the "-helix", "-helix-angles", and "-axis" arguments.)

-in coords.raw

The "-in" argument allows you to specify the name of a file with coordinates, instead of reading the coordinates from the standard input. (On rare occasions, such as while debugging, this can be more convenient. Most people will have no use for this feature.)

Examples:

  1. Make a simple polymer, adding "@bond:Backbone" type bonds between "$atom:c2" from each monomer with "$atom:c1" from the next monomer.
   genpoly_lt.py -bond Backbone c2 c1 < crds.raw > poly.lt
  1. Make a circular twisted double-stranded DNA model, treating each base-pair as a monomer, and connecting each base-pair monomer with 2 bonds with the next base-pair. This is done using 2 "-bond" commands connecting the "O3p_a" atom with the "P_a" atom (in strand A), and the "P_b" atom with the "O3p_b" atom (from the opposite strand, B).
   genpoly_lt.py -circular yes -helix 34.2857 \
                 -header 'import "basepair.lt"   #<--defines "BasePair"' \
                 -monomer-name "BasePair" \
                 -polymer-name "Plasmid" \
                 -bond Backbone O3p_a   P_a \
                 -bond Backbone P_b   O3p_b \
                 < dna_basepair_CM_coords.raw \
                 > chromosome.lt

If you want to control the sequence of the polymer, replace the "-monomer-name" argument with "-sequence sequence.txt".

Python API

It is possible to access the functionality of genpoly_lt.py from within python. To do that, you can create arrays of coordinates and python strings containing the names of the monomers in the polymer (see below). Then use GenPoly.WriteLTFile() create a file in MOLTEMPLATE (.LT) format. (If you prefer, the text in that file can be redirected to a python string using StringIO. In this way it is not necessary to read or write files to the file system.)

However, (as you can probably tell) making this possible within python was an afterthought. Currently, the easy way to do this is to pass the same command line arguments (described above) to GenPoly.ParseArgs() Then invoke the GenPoly.WriteLTFile() function.

However you can also edit the data members of the GenPoly object directly after it is created, instead of using the command-line arguments. The example below demonstrates how to specify the coordinates (coords_multi) and the names of the monomers (name_sequence_multi) so that you don't have specify this information in the argument list (or read any files from the file system). Perhaps in the future, I will clean this up.)*

class GenPoly
    """
    Read coordinates from a file, and generate a list of \"new\" commands
    in moltemplate format with the position of each monomer located
    at these positions, oriented appropriately, with bonds (and angles,
    dihedrals, etc...) connecting successive monomers together.
    By default, only a single polymer is created.
    However this class can create multiple polymers of different lengths.
    The list of coordinates for each polymer are saved separately within
    the "self.coords_multi" member.
    """

Usage example inside python

import math
import numpy as np
import moltemplate

N = 4
# Generate a zig-zag curve containing N points
x_orig = np.array([[i, 0.5*(i%2), 0.0] for i in range(0,N)])

# It's a good idea to generate a smoother version of this curve
# with 21 different positions along the curve (21=number of monomers).
x_new = moltemplate.interpolate_curve.ResampleCurve(x_orig, 21, 0.5)

# Optional:
# We want the spacing between monomers to be 0.332nm per monomer.
x_new *= 0.332 / ((math.sqrt(1+0.5**2)*len(x_orig)) / (len(x_new)-1))

# Now use genpoly_lt.GenPoly to generate an LT file describing
# a coarse-grained DNA molecule placed at all of these locations (x_new).
# (Note: Since there is only one polymer, the "coords_multi"
#  and "name_sequence_multi" arguments contain only one list each.
#  More generally they could contain multiple lists, one for each
#  polymer in the system.  Don't worry about this detail.)

gp = moltemplate.genpoly_lt.GenPoly()
gp.coords_multi = [x_new]

# Now specify the identity of each monomer in the polymer
# (In this case each "monomer" is a DNA base-pair, with names like "AT", "GC".)
gp.name_sequence_multi =[['AT', 'CG', 'GC', 'TA', 'AT', 'CG', 'GC', 'TA',
                          'AT', 'CG', 'GC', 'TA', 'AT', 'CG', 'GC', 'TA',
                          'AT', 'CG', 'GC', 'TA', 'AT']]

# The remaining settings are parsed from an argument list
gp.ParseArgs(['-helix', '34.2857',
              '-bond', 'Backbone', 'f', 'f',
              '-bond', 'Backbone', 'r', 'r',
              '-polymer-name', 'DNA_snippet',
              '-inherits', 'OXDNA2',
              '-header', 'import oxdna2.lt',
              '-circular', 'no'])

# Generate an .LT file and write it to the terminal (sys.stdout)
import sys
gp.WriteLTFile(sys.stdout)