allpy: lib/block.py annotate

allpy

annotate lib/block.py @ 129:a6d287f9c901

broken: lib::block::geometrical_cores returns list of blocks added new parameters to geometrical_core test passed

author	boris <bnagaev@gmail.com>
date	Sun, 24 Oct 2010 00:26:18 +0400
parents	3e69e4b38017
children	89de20cda19c

rev	line source
BurkovBA@0	1 #!usr/bin/python
BurkovBA@0	2
BurkovBA@0	3 import sys
BurkovBA@0	4
BurkovBA@0	5 import project
BurkovBA@0	6 import sequence
BurkovBA@0	7 import monomer
bnagaev@116	8 import config
bnagaev@121	9 from graph import Graph
bnagaev@129	10 from copy import copy
BurkovBA@0	11
BurkovBA@0	12 class Block(object):
BurkovBA@0	13 """
BurkovBA@0	14 Mandatory data:
BurkovBA@0	15 * self.project -- project object, which the block belongs to
BurkovBA@1	16 * self.sequences - set of sequence objects that contain monomers
BurkovBA@0	17 and/or gaps, that constitute the block
bnagaev@115	18 * self.positions -- sorted list of positions of the project.alignment that
BurkovBA@1	19 are included in the block
bnagaev@116	20
BurkovBA@0	21 How to create a new block:
BurkovBA@0	22 >>> import project
BurkovBA@0	23 >>> import block
BurkovBA@0	24 >>> proj = project.Project(open("test.fasta"))
bnagaev@114	25 >>> block1 = block.Block(proj)
BurkovBA@0	26 """
BurkovBA@0	27
bnagaev@112	28 def __init__(self, project, sequences=None, positions=None):
bnagaev@112	29 """
bnagaev@112	30 Builds new block from project
bnagaev@112	31 if sequences==None, all sequences are used
bnagaev@112	32 if positions==None, all positions are used
bnagaev@112	33 """
bnagaev@112	34 if sequences == None:
bnagaev@126	35 sequences = list(project.sequences)
bnagaev@112	36 if positions == None:
bnagaev@112	37 positions = range(len(project))
BurkovBA@73	38 self.project = project
BurkovBA@73	39 self.sequences = sequences
BurkovBA@73	40 self.positions = positions
BurkovBA@0	41
bnagaev@112	42 def save_fasta(self, out_file, long_line=60):
bnagaev@112	43 """
bnagaev@112	44 Saves alignment to given file in fasta-format
bnagaev@112	45 Splits long lines to substrings of length=long_line
bnagaev@112	46 To prevent this, set long_line=None
bnagaev@112	47
BurkovBA@0	48 No changes in the names, descriptions or order of the sequences
BurkovBA@0	49 are made.
BurkovBA@0	50 """
BurkovBA@0	51 for sequence in self.sequences:
bnagaev@112	52 out_file.write(">%(name)s %(description)s \n" % sequence.__dict__)
bnagaev@113	53 alignment_monomers = self.project.alignment[sequence]
bnagaev@115	54 block_monomers = [alignment_monomers[i] for i in self.positions]
bnagaev@113	55 string = ''.join([m.type.code1 if m else '-' for m in block_monomers])
bnagaev@112	56 if long_line:
bnagaev@112	57 for i in range(0, len(string) // long_line + 1):
bnagaev@112	58 out_file.write("%s \n" % string[ilong_line : ilong_line + long_line])
bnagaev@112	59 else:
bnagaev@112	60 out_file.write("%s \n" % string)
bnagaev@116	61
bnagaev@127	62 def geometrical_cores(self, max_delta=config.delta,
bnagaev@129	63 timeout=config.timeout, minsize=config.minsize,
bnagaev@129	64 ac_new_atoms=config.ac_new_atoms,
bnagaev@129	65 ac_count=config.ac_count):
bnagaev@116	66 """
bnagaev@129	67 returns length-sorted list of blocks, representing GCs
bnagaev@126	68
bnagaev@129	69 max_delta -- threshold of distance spreading
bnagaev@129	70 timeout -- Bron-Kerbosh timeout (then fast O(n ln n) algorithm)
bnagaev@129	71 minsize -- min size of each core
bnagaev@129	72 ac_new_atoms -- min part or new atoms in new alternative core
bnagaev@129	73 current GC is compared with each of already selected GCs
bnagaev@129	74 if difference is less then ac_new_atoms, current GC is skipped
bnagaev@129	75 difference = part of new atoms in current core
bnagaev@129	76 ac_count -- max number of cores (including main core)
bnagaev@120	77
bnagaev@120	78 If more than one pdb chain for some sequence provided, consider all of them
bnagaev@120	79 cost is calculated as 1 / (delta + 1)
bnagaev@120	80 delta in [0, +inf) => cost in (0, 1]
bnagaev@116	81 """
bnagaev@117	82 nodes = self.positions
bnagaev@117	83 lines = {}
bnagaev@116	84 for i in self.positions:
bnagaev@116	85 for j in self.positions:
bnagaev@117	86 if i < j:
bnagaev@120	87 distances = []
bnagaev@120	88 for sequence in self.sequences:
bnagaev@120	89 for chain in sequence.pdb_chains:
bnagaev@120	90 m1 = self.project.alignment[sequence][i]
bnagaev@120	91 m2 = self.project.alignment[sequence][j]
bnagaev@122	92 if m1 and m2:
bnagaev@122	93 ca1 = m1.pdb_residues[chain]['CA']
bnagaev@122	94 ca2 = m2.pdb_residues[chain]['CA']
bnagaev@122	95 d = ca1 - ca2 # Bio.PDB feature
bnagaev@122	96 distances.append(d)
bnagaev@122	97 if len(distances) >= 2:
bnagaev@122	98 delta = max(distances) - min(distances)
bnagaev@122	99 if delta <= max_delta:
bnagaev@122	100 lines[Graph.line(i, j)] = 1.0 / (1.0 + max_delta)
bnagaev@120	101 graph = Graph(nodes, lines)
bnagaev@129	102 cliques = graph.cliques(timeout=timeout, minsize=minsize)
bnagaev@129	103 GCs = []
bnagaev@129	104 for clique in cliques:
bnagaev@129	105 for GC in GCs:
bnagaev@129	106 if len(clique - set(GC.positions)) < ac_new_atoms * len(clique):
bnagaev@129	107 break
bnagaev@129	108 else:
bnagaev@129	109 GCs.append(Block(self.project, copy(self.sequences), clique))
bnagaev@129	110 if len(GCs) >= ac_count:
bnagaev@129	111 break
bnagaev@129	112 return GCs
bnagaev@123	113
bnagaev@128	114 def xstring(self, x='X'):
bnagaev@123	115 """
bnagaev@123	116 Returns string consisting of '-' and chars x at self.positions
bnagaev@123	117 Length of returning string = length of project
bnagaev@123	118 """
bnagaev@123	119 monomers = [False] * len(self.project)
bnagaev@123	120 for i in self.positions:
bnagaev@123	121 monomers[i] = True
bnagaev@123	122 return ''.join([x if m else '-' for m in monomers])