From PyMOLWiki
Revision as of 13:00, 8 November 2011 by Inchoate (talk | contribs) (The Code)
Jump to: navigation, search

Overview & Motivation

Anyone ever give you a protein and then say, find the sequence "FLVEW"? Well, this script will find any string or regular expression in a given object and return that selection for you. Here's an example,

# fetch two sugar-binding PDB
fetch 1tvn
# Now, find FATEW in 1tvn, similarly
findSeq FATEW, 1tvn
# lower-case works, too
findSeq fatew, 1tvn
# how about a regular expression?
findSeq F.*W, 1tvn

# Find the regular expression:
#  ..H[TA]LVWH
# in the few proteins loaded.
# I then showed them as sticks and colored them to highlight matched AAs
for x in cmd.get_names(): seqFinder( "..H[TA]LVWH", x, "sele_"+x, firstOnly=1 )
Red residues were those matching the regular expression '..H[TA]LVWH'.


I built this to be rather flexible. You call it as:

findSeq needle, haystack[, selName[, het[, firstOnly ]]]

where the options are:

needle the sequence of amino acids to find. Should be a string of one letter amino acid abbreviations. Can also be a string-style regular expression (eg. FW.*QQ).
haystack the PyMOL object or selection in which to search
selName the name of the returned selection. If you leave this blank, it'll be foundSeqXYZ where XYZ is some random integer (eg. foundSeq1435); if you supply sele then the usual PyMOL (sele) is used; and, finally, if it's anything else, then that will be used verbatim. Defaults to foundSeqXYZ so as not to overwrite any selections you might have in sele.
het 0/1 -- if 0 then heteroatoms are not considered; if 1 then they are; defaults to 0
firstOnly 0/1 -- if 0 then all matches are selected and returned; if 1 then only the first is returned

The Code

# -*- coding: utf-8 -*-
# --
# Finds given amino acids in a protein & returns the selection.
import re
import types
import random

import pymol
from pymol import stored, cmd

def checkParams(needle,haystack,selName,het,firstOnly):
	This is just a helper function for checking the user input
	# check Needle
	if len(needle)==0 or type(needle)!=types.StringType:
		print "Error: Please provide a string 'needle' to search for."
		print "Error: For help type 'help motifFinder'."
		return False
	# check Haystack
	if len(haystack)==0 or type(haystack)!=types.StringType:
		print "Error: Please provide valid PyMOL object or selection name"
		print "Error: in which to search."
		print "Error: For help type 'help motifFinder'."
		return False

	# check het
		het = bool(int(het))
	except ValueError:
		print "Error: The 'het' parameter was not 0 or 1."
		return False

	# check first Only
		firstOnly = bool(int(het))
	except ValueError:
		print "Error: The 'firstOnly' parameter was not 0 or 1."
		return False

	# check selName
	if type(selName)!=types.StringType:
		print "Error: selName was not a string."
		return False

	return True

def seqFinder(needle, haystack, selName=None, het=0, firstOnly=0):
	Given a sequence/regex to find, select those
	matching amino acids in the protein.

	seqFinder needle, haystack[, selName[, het[, firstOnly]]]

	needle (string)
			the sequence of amino acids to match and select
			in the haystack.  This can be a sequence of amino
			acids, or a string-style regular expression.  See
	hastack (string or PyMOL selection)
			name of the PyMOL object/selection in which 
			to find the needle.

	selName (string; defaults to None)
			This is the name of the selection to return.  If selName
			is left blank (None), then the selection name will be
			foundSeqXYZ where XYZ is some random number; if selName is
			"sele" the usual PyMOL "(sele)" will be used; and, lastly,
			if selName is anything else, that name will be used verbatim.

	het (0 or 1; defaults to 0)
			This boolean flag allows (1) or disallows (0) heteroatoms
			from being considered.

	firstOnly (0 or 1; defaults to 0)
			Subsequences or motifs might be repeated, this controls how we
			consider multiple matches.  If firstOnly is False (0) then we return
			all found subsequences; if firstOnly is True (1), then we just return
			the first found sequence.

	a newly created selection with the atoms you sought.  If there are
	more than two contiguous regions, then a newly created group is
	returned with each contiguous segment its own selection.

	# find SPVI in 1h12, foundSeqXYZ as return name
	seqFinder SPVI, 1h12
	# find FATEW and make it (sele).
	seqFinder FATEW, 1g01, sele
	# find the regular expression GMS.*QWY in 1a3h
	# and put the return value in (sele).
	fetch 1a3h
	# this ends up finding the sequence, GMSSHGLQWY
	seqFinder GMS.*QWY, 1a3h, sele

	Assumes we're using the ONE LETTER amino acid abbreviations.

	Jason Vertrees, 2009.

	# input checking
	if not checkParams(needle, haystack, selName, het, firstOnly):
		print "There was an error with a parameter.  Please see"
		print "the above error message for how to fix it."
		return None

	one_letter ={'VAL':'V', 'ILE':'I', 'LEU':'L', 'GLU':'E', 'GLN':'Q', \
	'ASP':'D', 'ASN':'N', 'HIS':'H', 'TRP':'W', 'PHE':'F', 'TYR':'Y',    \
	'ARG':'R', 'LYS':'K', 'SER':'S', 'THR':'T', 'MET':'M', 'ALA':'A',    \
	'GLY':'G', 'PRO':'P', 'CYS':'C'}
	# remove hetero atoms (waters/ligands/etc) from consideration?
	if het: "__h", "br. " + haystack )
	else: "__h", "br. " + haystack + " and not het" )
	# get the AAs in the haystack
	aaDict = { 'aaList' : [] }
	cmd.iterate("(name ca) and __h","aaList.append((resi,resn))",space=aaDict)
	IDs = map( lambda x: int(x[0]), aaDict['aaList'] )
	AAs = ''.join(map( lambda x: one_letter[x[1]], aaDict['aaList'] ))

	reNeedle = re.compile( needle.upper() )
	it = reNeedle.finditer( AAs )

	# set the name of the selection to return.
	if selName == None:
		rSelName = "foundSeq" + str(random.randint(0,32000))
	elif selName == "sele":
		rSelName = "sele"
		rSelName = selName

	# make an empty selection to which we add residues rSelName, 'None')

	for i in it:
		(start,stop) = i.span() rSelName, rSelName + " __h and i. " + str(IDs[start]) + "-" + str(IDs[stop-1]))
		if int(firstOnly):

	return rSelName

cmd.extend("findSeq", seqFinder )