From PyMOLWiki
Revision as of 18:11, 18 July 2009 by Shiven (talk | contribs) (Examples: Formatting)
Jump to: navigation, search


This script allows you to color two structures by Root Mean Square Deviation (RMSD), as calculated by PyMol's internal Rms_Cur command. The pairwise, C-alpha, RMSD is calculated only between those residues which are first used by PyMol's internal Super command to superpose the two structures. The RMSD values are stored as B-factors of these residues, which are colored by a rainbow color spectrum, with blue specifying the minimum pairwise RMSD and red indicating the maximum. Residues NOT used by Super for superposition, and hence for RMSD calculation, are colored white.


Please use this script with the option doPretty=T to gain the most benefit. Additionally, setting Cartoon representation to loop, with the command:

 cartoon loop

may help ease comparisons. Do keep in mind, all original B-factors values are overwritten!


# example #1
colorByRMSD 1cbs, 1hmt, doAlign=T, doPretty=T
# example #2
colorByRMSD 1eaz, 1fao, doAlign=T, doPretty=T
--- ColorByRMSD: RMSD based coloring --- 
Authors : Shivender Shandilya; Jason Vertrees
Program : ColorByRMSD
Date    : July 2009
Version : 0.1.1
Mail    :
Keywords: color rms rmsd colorbyrms colorbyrmsd
 This email from Warren -
 DeLano, W.L. The PyMOL Molecular Graphics System (2002) DeLano Scientific, San Carlos, CA, USA.
import pymol
import cmd
from pymol import stored
def strTrue(p):
    return p[0].upper() == "T"
# The main function that assigns current RMSD as the new B-factor
def rmsUpdateB(objA, alnAri, objB, alnBri):
    for x in range(len(alnAri)):
        s1 = objA + " and n. CA and i. " + alnAri[x]
        s2 = objB + " and n. CA and i. " + alnBri[x]
        rmsd = cmd.rms_cur(s1, s2, matchmaker=4)
        cmd.alter( s1, "b = " + str(rmsd))
        cmd.alter( s2, "b = " + str(rmsd))
    cmd.sort(objA); cmd.sort(objB)
def colorByRMSD(objSel1, objSel2, doAlign="True", doPretty=None):
    colorByRMSD -- align two structures and show the structural deviations
                   in color to more easily see variable regions.
        objSel1 (valid PyMOL object or selection)
            The first object to align.  
        objSel2 (valid PyMOL object or selection)
            The second object to align
        doAlign (boolean, either True or False)
            Should this script align your proteins or just leave them as is?
            If doAlign=True then your original proteins are aligned.
            If False, then they are not. Regardless, the B-factors are changed.
            DEFAULT: True
        doPretty (boolean, either True or False)
            If doPretty=True then a simple representation is created to
            highlight the differences.  If False, then no changes are made.
            DEFAULT: False
        Modifies the B-factor columns in your original structures.
    # First create backup copies; names starting with __ (underscores) are
    # normally hidden by PyMOL
    tObj1, tObj2, aln = "__tempObj1", "__tempObj2", "__aln"
    if strTrue(doAlign):
        # perform the alignment
        cmd.create( tObj1, objSel1 )
        cmd.create( tObj2, objSel2 )
        cmd.super( tObj1, tObj2, object=aln )
        cmd.matrix_copy(tObj1, objSel1)
        # perform the alignment
        cmd.create( tObj1, objSel1 )
        cmd.create( tObj2, objSel2 )
        cmd.super( tObj1, tObj2, object=aln )
    # Modify the B-factor columns of the original objects,
    # in order to identify the residues NOT used for alignment, later on
    cmd.alter( objSel1 + " or " + objSel2, "b=-10")
    cmd.alter( tObj1 + " or " + tObj2, "chain='A'")
    cmd.alter( tObj1 + " or " + tObj2, "segi='A'")
    # Update pymol internal representations; one of these should do the trick
    cmd.refresh(); cmd.rebuild(); cmd.sort(tObj1); cmd.sort(tObj2)
    #  Create lists for storage
    stored.alnAres, stored.alnBres = [], []
    #  Get the residue identifiers from the alignment object "aln"
    cmd.iterate(tObj1 + " and n. CA and " + aln, "stored.alnAres.append(resi)")
    cmd.iterate(tObj2 + " and n. CA and " + aln, "stored.alnBres.append(resi)")
    # Change the B-factors for EACH object
    # Store the NEW B-factors
    stored.alnAnb, stored.alnBnb = [], []
    cmd.iterate(tObj1 + " and n. CA and " + aln, "stored.alnAnb.append(b)" )
    cmd.iterate(tObj2 + " and n. CA and " + aln, "stored.alnBnb.append(b)" )
    # Get rid of all intermediate objects and clean up
    # Assign the just stored NEW B-factors to the original objects
    for x in range(len(stored.alnAres)):
        cmd.alter(objSel1 + " and n. CA and i. " + str(stored.alnAres[x]), "b = " + str(stored.alnAnb[x]))
    for x in range(len(stored.alnBres)):
        cmd.alter(objSel2 + " and n. CA and i. " + str(stored.alnBres[x]), "b = " + str(stored.alnBnb[x]))
    cmd.rebuild(); cmd.refresh(); cmd.sort(objSel1); cmd.sort(objSel2)

    # Provide some useful information
    stored.allRMSDval = []
    stored.allRMSDval = stored.alnAnb + stored.alnBnb
    print "\nColorByRMSD completed successfully."
    print "The MINIMUM RMSD value is: "+str(min(stored.allRMSDval))
    print "The MAXIMUM RMSD value is: "+str(max(stored.allRMSDval))

    if doPretty!=None:
        # Showcase what we did
        cmd.show_as("cartoon", objSel1 + " or " + objSel2)
        # Select the residues not used for alignment; they still have their B-factors as "-10""notUsedForAln", "b < 0")
	 # White-wash the residues not used for alignment
	 cmd.color("white", "notUsedForAln")
	 # Color the residues used for alignment according to their B-factors (RMSD values)
        cmd.spectrum("b", 'rainbow',  "((" + objSel1 + " and n. CA) or (n. CA and " + objSel2 +" )) and not notUsedForAln")
        # Delete the selection of atoms not used for alignment
        # If you would like to keep this selection intact,
        # just comment "cmd.delete" line and
        # uncomment the "cmd.disable" line below.
        # cmd.disable("notUsedForAln") 

        print "\nObjects are now colored by C-alpha RMS deviation."
        print "All residues with RMSD values greater than the maximum are colored white..."
cmd.extend("colorByRMSD", colorByRMSD)