allpy: 56bea23d3c73 lib/block.py

allpy

view lib/block.py @ 184:56bea23d3c73

repeats: draft is completed, not tested TODO: take not exactly noticed repeats, but also part of environment (for muscle alignment)

author	boris (netbook) <bnagaev@gmail.com>
date	Wed, 10 Nov 2010 12:01:26 +0300
parents	45573ee48844
children

line source

1 #!usr/bin/python

3 import sys

5 import alignment

6 import sequence

7 import monomer

8 import config

9 from graph import Graph

10 from Bio.PDB import Superimposer

11 from tempfile import NamedTemporaryFile

12 import os

13 from fasta import save_fasta

15 class Block(object):

16 """ Block of alignment

18 Mandatory data:

19 * self.alignment -- alignment object, which the block belongs to

20 * self.sequences - set of sequence objects that contain monomers

21 and/or gaps, that constitute the block

22 * self.positions -- list of positions of the alignment.body that

23 are included in the block; position[i+1] is always to the right from position[i]

25 Don't change self.sequences -- it may be a link to other block.sequences

27 How to create a new block:

28 >>> import alignment

29 >>> import block

30 >>> proj = alignment.Alignment(open("test.fasta"))

31 >>> block1 = block.Block(proj)

32 """

34 def __init__(self, alignment, sequences=None, positions=None):

35 """ Builds new block from alignment

37 if sequences==None, all sequences are used

38 if positions==None, all positions are used

39 """

40 if sequences == None:

41 sequences = set(alignment.sequences) # copy

42 if positions == None:

43 positions = range(len(alignment))

44 self.alignment = alignment

45 self.sequences = sequences

46 self.positions = positions

48 def save_fasta(self, out_file, long_line=70, gap='-'):

49 """ Saves alignment to given file in fasta-format

51 No changes in the names, descriptions or order of the sequences

52 are made.

53 """

54 for sequence in self.sequences:

55 alignment_monomers = self.alignment.body[sequence]

56 block_monomers = [alignment_monomers[i] for i in self.positions]

57 string = ''.join([m.type.code1 if m else '-' for m in block_monomers])

58 save_fasta(out_file, string, sequence.name, sequence.description, long_line)

60 def geometrical_cores(self, max_delta=config.delta,

61 timeout=config.timeout, minsize=config.minsize,

62 ac_new_atoms=config.ac_new_atoms,

63 ac_count=config.ac_count):

64 """ Returns length-sorted list of blocks, representing GCs

66 max_delta -- threshold of distance spreading

67 timeout -- Bron-Kerbosh timeout (then fast O(n ln n) algorithm)

68 minsize -- min size of each core

69 ac_new_atoms -- min part or new atoms in new alternative core

70 current GC is compared with each of already selected GCs

71 if difference is less then ac_new_atoms, current GC is skipped

72 difference = part of new atoms in current core

73 ac_count -- max number of cores (including main core)

74 -1 means infinity

75 If more than one pdb chain for some sequence provided, consider all of them

76 cost is calculated as 1 / (delta + 1)

77 delta in [0, +inf) => cost in (0, 1]

78 """

79 nodes = self.positions

80 lines = {}

81 for i in self.positions:

82 for j in self.positions:

83 if i < j:

84 distances = []

85 for sequence in self.sequences:

86 for chain in sequence.pdb_chains:

87 m1 = self.alignment.body[sequence][i]

88 m2 = self.alignment.body[sequence][j]

89 if m1 and m2:

90 r1 = sequence.pdb_residues[chain][m1]

91 r2 = sequence.pdb_residues[chain][m2]

92 ca1 = r1['CA']

93 ca2 = r2['CA']

94 d = ca1 - ca2 # Bio.PDB feature

95 distances.append(d)

96 if len(distances) >= 2:

97 delta = max(distances) - min(distances)

98 if delta <= max_delta:

99 lines[Graph.line(i, j)] = 1.0 / (1.0 + max_delta)

100 graph = Graph(nodes, lines)

101 cliques = graph.cliques(timeout=timeout, minsize=minsize)

102 GCs = []

103 for clique in cliques:

104 for GC in GCs:

105 if len(clique - set(GC.positions)) < ac_new_atoms * len(clique):

106 break

107 else:

108 GCs.append(Block(self.alignment, self.sequences, clique))

109 if ac_count != -1 and len(GCs) >= ac_count:

110 break

111 return GCs

112

113 def xstring(self, x='X', gap='-'):

114 """ Returns string consisting of gap chars and chars x at self.positions

115

116 Length of returning string = length of alignment

117 """

118 monomers = [False] * len(self.alignment)

119 for i in self.positions:

120 monomers[i] = True

121 return ''.join([x if m else gap for m in monomers])

122

123 def save_xstring(self, out_file, name, description='', x='X', gap='-', long_line=70):

124 """ Save xstring and name in fasta format """

125 save_fasta(out_file, self.xstring(x=x, gap=gap), name, description, long_line)

126

127 def monomers(self, sequence):

128 """ Iterates monomers of this sequence from this block """

129 alignment_sequence = self.alignment.body[sequence]

130 return (alignment_sequence[i] for i in self.positions)

131

132 def ca_atoms(self, sequence, pdb_chain):

133 """ Iterates Ca-atom of monomers of this sequence from this block """

134 return (sequence.pdb_residues[pdb_chain][monomer] for monomer in self.monomers())

135

136 def sequences_chains(self):

137 """ Iterates pairs (sequence, chain) """

138 for sequence in self.alignment.sequences:

139 if sequence in self.sequences:

140 for chain in sequence.pdb_chains:

141 yield (sequence, chain)

142

143 def superimpose(self):

144 """ Superimpose all pdb_chains in this block """

145 sequences_chains = list(self.sequences_chains())

146 if len(sequences_chains) >= 1:

147 sup = Superimposer()

148 fixed_sequence, fixed_chain = sequences_chains.pop()

149 fixed_atoms = self.ca_atoms(fixed_sequence, fixed_chain)

150 for sequence, chain in sequences_chains:

151 moving_atoms = self.ca_atoms(sequence, chain)

152 sup.set_atoms(fixed_atoms, moving_atoms)

153 # Apply rotation/translation to the moving atoms

154 sup.apply(moving_atoms)

155

156 def pdb_save(self, out_file):

157 """ Save all sequences

158

159 Returns {(sequence, chain): CHAIN}

160 CHAIN is chain letter in new file

161 """

162 tmp_file = NamedTemporaryFile(delete=False)

163 tmp_file.close()

164

165 for sequence, chain in self.sequences_chains():

166 sequence.pdb_save(tmp_file.name, chain)

167 # TODO: read from tmp_file.name

168 # change CHAIN

169 # add to out_file

170

171 os.unlink(NamedTemporaryFile)

172