allpy: allpy/base.py annotate

allpy

annotate allpy/base.py @ 284:bee4d155f526

Added base.Sequence.from_fasta

author	Daniil Alexeyevsky <me.dendik@gmail.com>
date	Wed, 15 Dec 2010 23:44:20 +0300
parents	59320c160dae
children	cf6cdc3b7ec5

rev	line source
me@261	1 import sys
me@261	2 import os
me@262	3 import os.path
me@261	4 from tempfile import NamedTemporaryFile
me@262	5 import urllib2
me@261	6
me@261	7 import config
me@284	8 import fasta
me@261	9 from graph import Graph
me@262	10 from Bio.PDB.DSSP import make_dssp_dict
me@260	11 import data.codes
me@260	12
me@260	13 class MonomerType(object):
me@260	14 """Class of monomer types.
me@260	15
me@260	16 Each MonomerType object represents a known monomer type, e.g. Valine,
me@260	17 and is referenced to by each instance of monomer in a given sequence.
me@260	18
me@260	19 - `name`: full name of monomer type
me@260	20 - `code1`: one-letter code
me@260	21 - `code3`: three-letter code
me@260	22 - `is_modified`: either of True or False
me@260	23
me@260	24 class atributes:
me@260	25
me@260	26 - `by_code1`: a mapping from one-letter code to MonomerType object
me@260	27 - `by_code3`: a mapping from three-letter code to MonomerType object
me@260	28 - `by_name`: a mapping from monomer name to MonomerType object
me@260	29 - `instance_type`: class of Monomer objects to use when creating new
me@260	30 objects; this must be redefined in descendent classes
me@260	31
me@260	32 All of the class attributes MUST be redefined when subclassing.
me@260	33 """
me@260	34
me@260	35 by_code1 = {}
me@260	36 by_code3 = {}
me@260	37 by_name = {}
me@260	38 instance_type = None
me@260	39
me@260	40 def __init__(self, name="", code1="", code3="", is_modified=False):
me@260	41 self.name = name.capitalize()
me@260	42 self.code1 = code1.upper()
me@260	43 self.code3 = code3.upper()
me@260	44 self.is_modified = bool(is_modified)
me@260	45 if not is_modified:
me@260	46 self.by_code1[self.code1] = self
me@260	47 self.by_code3[code3] = self
me@260	48 self.by_name[name] = self
me@260	49 # We duplicate distinguished long names into MonomerType itself,
me@260	50 # so that we can use MonomerType.from_code3 to create the relevant
me@260	51 # type of monomer.
me@260	52 MonomerType.by_code3[code3] = self
me@260	53 MonomerType.by_name[name] = self
me@260	54
me@260	55 @classmethod
me@260	56 def _initialize(cls, type_letter, codes=data.codes.codes):
me@260	57 """Create all relevant instances of MonomerType.
me@260	58
me@260	59 `type_letter` is either of:
me@260	60
me@260	61 - 'p' for protein
me@260	62 - 'd' for DNA
me@260	63 - 'r' for RNA
me@260	64
me@260	65 `codes` is a table of monomer codes
me@260	66 """
me@260	67 for type, code1, is_modified, code3, name in codes:
me@260	68 if type == type_letter:
me@260	69 cls(name, code1, code3, is_modified)
me@260	70
me@260	71 @classmethod
me@260	72 def from_code1(cls, code1):
me@260	73 """Return monomer type by one-letter code."""
me@260	74 return cls.by_code1[code1.upper()]
me@260	75
me@260	76 @classmethod
me@260	77 def from_code3(cls, code3):
me@260	78 """Return monomer type by three-letter code."""
me@260	79 return cls.by_code3[code3.upper()]
me@260	80
me@260	81 @classmethod
me@260	82 def from_name(cls, name):
me@260	83 """Return monomer type by name."""
me@260	84 return cls.by_name[name.capitalize()]
me@260	85
me@260	86 def instance(self):
me@260	87 """Create a new monomer of given type."""
me@260	88 return self.instance_type(self)
me@260	89
me@260	90 def __eq__(self, other):
me@260	91 if hasattr(other, "type"):
me@260	92 return self is other.type
me@260	93 return self is other
me@260	94
me@260	95 class Monomer(object):
me@260	96 """Monomer object.
me@260	97
me@260	98 attributes:
me@260	99
me@260	100 - `type`: type of monomer (a MonomerType object)
me@260	101
me@282	102 class attributes:
me@282	103
me@282	104 - `monomer_type`: either MonomerType or one of it's subclasses, it is used
me@282	105 when creating new monomers. It SHOULD be redefined when subclassing
me@282	106 Monomer.
me@260	107 """
me@260	108 monomer_type = MonomerType
me@260	109
me@260	110 def __init__(self, type):
me@260	111 self.type = type
me@260	112
me@260	113 @classmethod
me@260	114 def from_code1(cls, code1):
me@260	115 return cls(cls.monomer_type.by_code1[code1.upper()])
me@260	116
me@260	117 @classmethod
me@260	118 def from_code3(cls, code3):
me@260	119 return cls(cls.monomer_type.by_code3[code3.upper()])
me@260	120
me@260	121 @classmethod
me@260	122 def from_name(cls, name):
me@260	123 return cls(cls.monomer_type.by_name[name.capitalize()])
me@260	124
me@260	125 def __eq__(self, other):
me@260	126 if hasattr(other, "type"):
me@260	127 return self.type is other.type
me@260	128 return self.type is other
bnagaev@239	129
bnagaev@239	130 class Sequence(list):
me@274	131 """Sequence of Monomers.
bnagaev@243	132
me@274	133 This behaves like list of monomer objects. In addition to standard list
me@274	134 behaviour, Sequence has the following attributes:
me@270	135
me@274	136 * name -- str with the name of the sequence
me@274	137 * description -- str with description of the sequence
me@274	138 * source -- str denoting source of the sequence
me@266	139
me@274	140 Any of them may be empty (i.e. hold empty string)
me@275	141
me@275	142 Class attributes:
me@282	143
me@275	144 * monomer_type -- type of monomers in sequence, must be redefined when
me@275	145 subclassing
me@274	146 """
me@270	147
me@275	148 monomer_type = Monomer
me@270	149
me@275	150 name = ''
me@275	151 description = ''
me@275	152 source = ''
me@275	153
me@275	154 def __init__(self, sequence=[], name=None, description=None, source=None):
me@275	155 super(Sequence, self).__init__(sequence)
me@275	156 if hasattr(sequence, 'name'):
me@275	157 vars(self).update(vars(sequence))
me@275	158 if name:
me@275	159 self.name = name
me@275	160 if description:
me@275	161 self.description = description
me@275	162 if source:
me@275	163 self.source = source
me@270	164
me@262	165 def __str__(self):
me@275	166 """Returns sequence in one-letter code."""
me@275	167 return ''.join(monomer.code1 for monomer in self)
me@270	168
me@273	169 @classmethod
me@273	170 def from_string(cls, string, name='', description=''):
me@273	171 """Create sequences from string of one-letter codes."""
me@273	172 monomer = cls.monomer_type.from_code1
me@273	173 monomers = [monomer(letter) for letter in string]
me@273	174 return cls(monomers, name, description)
me@262	175
me@284	176 @classmethod
me@284	177 def from_fasta(cls, file):
me@284	178 """Read sequence from FASTA file.
me@284	179
me@284	180 File must contain exactly one sequence.
me@284	181 """
me@284	182 sequences = fasta.parse_file(file)
me@284	183 assert len(sequences) == 1
me@284	184 header = sequences.keys()[0]
me@284	185 name, _, description = header.partition(" ")
me@284	186 return cls(sequences[header], name, description, file.name)
me@284	187
bnagaev@249	188 class Alignment(dict):
me@282	189 """Alignment.
me@270	190
me@282	191 Behaves like a list of Columns.
bnagaev@249	192 """
bnagaev@249	193 # _sequences -- list of Sequence objects. Sequences don't contain gaps
bnagaev@249	194 # - see sequence.py module
bnagaev@249	195
bnagaev@249	196 def __init__(self, *args):
bnagaev@249	197 """overloaded constructor
bnagaev@249	198
me@282	199 Alignment()
me@282	200 new empty Alignment
bnagaev@249	201
me@282	202 Alignment(sequences, body)
me@282	203 new Alignment with sequences and body initialized from arguments
me@282	204
me@282	205 Alignment(fasta_file)
me@282	206 new Alignment, read body and sequences from fasta file
bnagaev@249	207 """
bnagaev@249	208 if len(args)>1:#overloaded constructor
bnagaev@249	209 self.sequences=args[0]
bnagaev@249	210 self.body=args[1]
bnagaev@249	211 elif len(args)==0:
bnagaev@249	212 self.sequences=[]
bnagaev@249	213 self.body={}
bnagaev@249	214 else:
bnagaev@249	215 self.sequences, self.body = Alignment.from_fasta(args[0])
bnagaev@249	216
bnagaev@249	217 def length(self):
bnagaev@249	218 """ Returns width, ie length of each sequence with gaps """
bnagaev@249	219 return max([len(line) for line in self.body.values()])
bnagaev@249	220
bnagaev@249	221 def height(self):
bnagaev@249	222 """ The number of sequences in alignment (it's thickness). """
bnagaev@249	223 return len(self.body)
bnagaev@249	224
bnagaev@249	225 def identity(self):
bnagaev@249	226 """ Calculate the identity of alignment positions for colouring.
bnagaev@249	227
bnagaev@249	228 For every (row, column) in alignment the percentage of the exactly
bnagaev@249	229 same residue in the same column in the alignment is calculated.
me@270	230 The data structure is just like the Alignment.body, but istead of
bnagaev@249	231 monomers it contains float percentages.
bnagaev@249	232 """
bnagaev@249	233 # Oh, God, that's awful! Absolutely not understandable.
bnagaev@249	234 # First, calculate percentages of amino acids in every column
bnagaev@249	235 contribution = 1.0 / len(self.sequences)
bnagaev@249	236 all_columns = []
bnagaev@249	237 for position in range(len(self)):
bnagaev@249	238 column_percentage = {}
bnagaev@249	239 for seq in self.body:
bnagaev@249	240 if self.body[seq][position] is not None:
bnagaev@249	241 aa = self.body[seq][position].code
bnagaev@249	242 else:
bnagaev@249	243 aa = None
bnagaev@249	244 if aa in allpy.data.amino_acids:
bnagaev@249	245 if aa in column_percentage.keys():
bnagaev@249	246 column_percentage[aa] += contribution
bnagaev@249	247 else:
bnagaev@249	248 column_percentage[aa] = contribution
bnagaev@249	249 all_columns.append(column_percentage)
bnagaev@249	250 # Second, map these percentages onto the alignment
bnagaev@249	251 self.identity_percentages = {}
bnagaev@249	252 for seq in self.sequences:
bnagaev@249	253 self.identity_percentages[seq] = []
bnagaev@249	254 for seq in self.identity_percentages:
bnagaev@249	255 line = self.identity_percentages[seq]
bnagaev@249	256 for position in range(len(self)):
bnagaev@249	257 if self.body[seq][position] is not None:
bnagaev@249	258 aa = self.body[seq][position].code
bnagaev@249	259 else:
bnagaev@249	260 aa = None
bnagaev@249	261 line.append(all_columns[position].get(aa))
bnagaev@249	262 return self.identity_percentages
bnagaev@249	263
me@277	264 @classmethod
me@277	265 def from_fasta(file):
bnagaev@249	266 """ Import data from fasta file
me@270	267
bnagaev@249	268 >>> import alignment
me@270	269 >>> sequences,body=alignment.Alignment.from_fasta(open("test.fasta"))
bnagaev@249	270 """
bnagaev@249	271 import re
bnagaev@249	272
bnagaev@249	273 sequences = []
bnagaev@249	274 body = {}
bnagaev@249	275
bnagaev@249	276 raw_sequences = file.read().split(">")
bnagaev@249	277 if len(raw_sequences) <= 1:
bnagaev@249	278 raise Exception("Wrong format of fasta-file %s" % file.name)
me@270	279
bnagaev@249	280 raw_sequences = raw_sequences[1:] #ignore everything before the first >
bnagaev@249	281 for raw in raw_sequences:
bnagaev@249	282 parsed_raw_sequence = raw.split("\n")
bnagaev@249	283 parsed_raw_sequence = [s.strip() for s in parsed_raw_sequence]
bnagaev@249	284 name_and_description = parsed_raw_sequence[0]
bnagaev@249	285 name_and_description = name_and_description.split(" ",1)
bnagaev@249	286 if len(name_and_description) == 2:
bnagaev@249	287 name, description = name_and_description
me@270	288 elif len(name_and_description) == 1:
bnagaev@249	289 #if there is description
bnagaev@249	290 name = name_and_description[0]
bnagaev@249	291 description = ''
bnagaev@249	292 else:
bnagaev@249	293 raise Exception("Wrong name of sequence %(name)$ fasta-file %(file)s" % \
bnagaev@249	294 {'name': name, 'file': file.name})
me@270	295
bnagaev@249	296 if len(parsed_raw_sequence) <= 1:
bnagaev@249	297 raise Exception("Wrong format of sequence %(name)$ fasta-file %(file)s" % \
bnagaev@249	298 {'name': name, 'file': file.name})
bnagaev@249	299 string = ""
bnagaev@249	300 for piece in parsed_raw_sequence[1:]:
bnagaev@249	301 piece_without_whitespace_chars = re.sub("\s", "", piece)
bnagaev@249	302 string += piece_without_whitespace_chars
bnagaev@249	303 monomers = [] #convert into Monomer objects
bnagaev@249	304 body_list = [] #create the respective list in body dict
bnagaev@249	305 for current_monomer in string:
bnagaev@249	306 if current_monomer not in ["-", ".", "~"]:
me@277	307 monomers.append(cls.monomer_type.from_code1(current_monomer))
bnagaev@249	308 body_list.append(monomers[-1])
bnagaev@249	309 else:
bnagaev@249	310 body_list.append(None)
bnagaev@249	311 s = sequence.Sequence(monomers, name, description)
bnagaev@249	312 sequences.append(s)
bnagaev@249	313 body[s] = body_list
bnagaev@249	314 return sequences, body
me@270	315
bnagaev@249	316 @staticmethod
bnagaev@249	317 def from_sequences(*sequences):
bnagaev@249	318 """ Constructs new alignment from sequences
me@270	319
me@270	320 Add None's to right end to make equal lengthes of alignment sequences
bnagaev@249	321 """
bnagaev@249	322 alignment = Alignment()
bnagaev@249	323 alignment.sequences = sequences
bnagaev@249	324 max_length = max(len(sequence) for sequence in sequences)
bnagaev@249	325 for sequence in sequences:
bnagaev@249	326 gaps_count = max_length - len(sequence)
bnagaev@249	327 alignment.body[sequence] = sequence.monomers + [None] * gaps_count
bnagaev@249	328 return alignment
me@270	329
bnagaev@249	330 def save_fasta(self, out_file, long_line=70, gap='-'):
bnagaev@249	331 """ Saves alignment to given file
me@270	332
bnagaev@249	333 Splits long lines to substrings of length=long_line
me@270	334 To prevent this, set long_line=None
bnagaev@249	335 """
bnagaev@249	336 block.Block(self).save_fasta(out_file, long_line=long_line, gap=gap)
me@270	337
bnagaev@249	338 def muscle_align(self):
bnagaev@249	339 """ Simple align ths alignment using sequences (muscle)
me@270	340
bnagaev@249	341 uses old Monomers and Sequences objects
bnagaev@249	342 """
bnagaev@249	343 tmp_file = NamedTemporaryFile(delete=False)
bnagaev@249	344 self.save_fasta(tmp_file)
bnagaev@249	345 tmp_file.close()
bnagaev@249	346 os.system("muscle -in %(tmp)s -out %(tmp)s" % {'tmp': tmp_file.name})
bnagaev@249	347 sequences, body = Alignment.from_fasta(open(tmp_file.name))
bnagaev@249	348 for sequence in self.sequences:
bnagaev@249	349 try:
bnagaev@249	350 new_sequence = [i for i in sequences if sequence==i][0]
bnagaev@249	351 except:
bnagaev@249	352 raise Exception("Align: Cann't find sequence %s in muscle output" % \
bnagaev@249	353 sequence.name)
bnagaev@249	354 old_monomers = iter(sequence.monomers)
bnagaev@249	355 self.body[sequence] = []
bnagaev@249	356 for monomer in body[new_sequence]:
bnagaev@249	357 if not monomer:
bnagaev@249	358 self.body[sequence].append(monomer)
bnagaev@249	359 else:
bnagaev@249	360 old_monomer = old_monomers.next()
bnagaev@249	361 if monomer != old_monomer:
bnagaev@249	362 raise Exception("Align: alignment errors")
bnagaev@249	363 self.body[sequence].append(old_monomer)
bnagaev@249	364 os.unlink(tmp_file.name)
me@270	365
bnagaev@249	366 def column(self, sequence=None, sequences=None, original=None):
bnagaev@249	367 """ returns list of columns of alignment
me@270	368
bnagaev@249	369 sequence or sequences:
me@282	370 * if sequence is given, then column is (original_monomer, monomer)
me@282	371 * if sequences is given, then column is (original_monomer, {sequence: monomer})
me@282	372 * if both of them are given, it is an error
me@282	373
bnagaev@249	374 original (Sequence type):
me@282	375 * if given, this filters only columns represented by original sequence
bnagaev@249	376 """
bnagaev@249	377 if sequence and sequences:
bnagaev@249	378 raise Exception("Wrong usage. read help")
bnagaev@249	379 indexes = dict([(v, k) for( k, v) in enumerate(self.sequences)])
bnagaev@249	380 alignment = self.body.items()
bnagaev@249	381 alignment.sort(key=lambda i: indexes[i[0]])
bnagaev@249	382 alignment = [monomers for seq, monomers in alignment]
bnagaev@249	383 for column in zip(*alignment):
bnagaev@249	384 if not original or column[indexes[original]]:
bnagaev@249	385 if sequence:
bnagaev@249	386 yield (column[indexes[original]], column[indexes[sequence]])
bnagaev@249	387 else:
me@270	388 yield (column[indexes[original]],
bnagaev@249	389 dict([(s, column[indexes[s]]) for s in sequences]))
me@270	390
bnagaev@249	391 def secstr(self, sequence, pdb_chain, gap='-'):
bnagaev@249	392 """ Returns string representing secondary structure """
bnagaev@249	393 return ''.join([
me@270	394 (sequence.pdb_secstr[pdb_chain][m] if sequence.secstr_has(pdb_chain, m) else gap)
bnagaev@249	395 for m in self.body[sequence]])
bnagaev@249	396
bnagaev@249	397 class Block(object):
me@261	398 """ Block of alignment
me@270	399
me@261	400 Mandatory data:
me@266	401
me@261	402 * self.alignment -- alignment object, which the block belongs to
me@261	403 * self.sequences - set of sequence objects that contain monomers
me@261	404 and/or gaps, that constitute the block
me@261	405 * self.positions -- list of positions of the alignment.body that
me@261	406 are included in the block; position[i+1] is always to the right from position[i]
me@270	407
me@261	408 Don't change self.sequences -- it may be a link to other block.sequences
me@270	409
me@261	410 How to create a new block:
me@282	411
me@261	412 >>> import alignment
me@261	413 >>> import block
me@261	414 >>> proj = alignment.Alignment(open("test.fasta"))
me@261	415 >>> block1 = block.Block(proj)
me@261	416 """
me@270	417
me@261	418 def __init__(self, alignment, sequences=None, positions=None):
me@261	419 """ Builds new block from alignment
me@270	420
me@261	421 if sequences==None, all sequences are used
me@261	422 if positions==None, all positions are used
me@261	423 """
me@261	424 if sequences == None:
me@261	425 sequences = set(alignment.sequences) # copy
me@261	426 if positions == None:
me@261	427 positions = range(len(alignment))
me@261	428 self.alignment = alignment
me@261	429 self.sequences = sequences
me@261	430 self.positions = positions
me@270	431
me@261	432 def save_fasta(self, out_file, long_line=70, gap='-'):
me@270	433 """ Saves alignment to given file in fasta-format
me@270	434
me@261	435 No changes in the names, descriptions or order of the sequences
me@261	436 are made.
me@261	437 """
me@261	438 for sequence in self.sequences:
me@261	439 alignment_monomers = self.alignment.body[sequence]
me@261	440 block_monomers = [alignment_monomers[i] for i in self.positions]
me@261	441 string = ''.join([m.type.code1 if m else '-' for m in block_monomers])
me@261	442 save_fasta(out_file, string, sequence.name, sequence.description, long_line)
me@270	443
me@270	444 def geometrical_cores(self, max_delta=config.delta,
me@270	445 timeout=config.timeout, minsize=config.minsize,
me@261	446 ac_new_atoms=config.ac_new_atoms,
me@261	447 ac_count=config.ac_count):
me@261	448 """ Returns length-sorted list of blocks, representing GCs
me@270	449
me@282	450 * max_delta -- threshold of distance spreading
me@282	451 * timeout -- Bron-Kerbosh timeout (then fast O(n ln n) algorithm)
me@282	452 * minsize -- min size of each core
me@282	453 * ac_new_atoms -- min part or new atoms in new alternative core
me@282	454 current GC is compared with each of already selected GCs if
me@282	455 difference is less then ac_new_atoms, current GC is skipped
me@261	456 difference = part of new atoms in current core
me@282	457 * ac_count -- max number of cores (including main core)
me@261	458 -1 means infinity
me@282	459
me@261	460 If more than one pdb chain for some sequence provided, consider all of them
me@270	461 cost is calculated as 1 / (delta + 1)
me@282	462
me@261	463 delta in [0, +inf) => cost in (0, 1]
me@261	464 """
me@261	465 nodes = self.positions
me@261	466 lines = {}
me@261	467 for i in self.positions:
me@261	468 for j in self.positions:
me@261	469 if i < j:
me@261	470 distances = []
me@261	471 for sequence in self.sequences:
me@261	472 for chain in sequence.pdb_chains:
me@261	473 m1 = self.alignment.body[sequence][i]
me@261	474 m2 = self.alignment.body[sequence][j]
me@261	475 if m1 and m2:
me@261	476 r1 = sequence.pdb_residues[chain][m1]
me@261	477 r2 = sequence.pdb_residues[chain][m2]
me@261	478 ca1 = r1['CA']
me@261	479 ca2 = r2['CA']
me@261	480 d = ca1 - ca2 # Bio.PDB feature
me@261	481 distances.append(d)
me@261	482 if len(distances) >= 2:
me@261	483 delta = max(distances) - min(distances)
me@261	484 if delta <= max_delta:
me@261	485 lines[Graph.line(i, j)] = 1.0 / (1.0 + max_delta)
me@261	486 graph = Graph(nodes, lines)
me@261	487 cliques = graph.cliques(timeout=timeout, minsize=minsize)
me@261	488 GCs = []
me@261	489 for clique in cliques:
me@261	490 for GC in GCs:
me@261	491 if len(clique - set(GC.positions)) < ac_new_atoms * len(clique):
me@261	492 break
me@261	493 else:
me@261	494 GCs.append(Block(self.alignment, self.sequences, clique))
me@261	495 if ac_count != -1 and len(GCs) >= ac_count:
me@261	496 break
me@261	497 return GCs
me@270	498
me@261	499 def xstring(self, x='X', gap='-'):
me@261	500 """ Returns string consisting of gap chars and chars x at self.positions
me@270	501
me@261	502 Length of returning string = length of alignment
me@261	503 """
me@261	504 monomers = [False] * len(self.alignment)
me@261	505 for i in self.positions:
me@261	506 monomers[i] = True
me@261	507 return ''.join([x if m else gap for m in monomers])
me@270	508
me@261	509 def save_xstring(self, out_file, name, description='', x='X', gap='-', long_line=70):
me@261	510 """ Save xstring and name in fasta format """
me@261	511 save_fasta(out_file, self.xstring(x=x, gap=gap), name, description, long_line)
me@270	512
me@261	513 def monomers(self, sequence):
me@261	514 """ Iterates monomers of this sequence from this block """
me@261	515 alignment_sequence = self.alignment.body[sequence]
me@261	516 return (alignment_sequence[i] for i in self.positions)
me@270	517
me@261	518 def ca_atoms(self, sequence, pdb_chain):
me@261	519 """ Iterates Ca-atom of monomers of this sequence from this block """
me@261	520 return (sequence.pdb_residues[pdb_chain][monomer] for monomer in self.monomers())
me@270	521
me@261	522 def sequences_chains(self):
me@261	523 """ Iterates pairs (sequence, chain) """
me@261	524 for sequence in self.alignment.sequences:
me@261	525 if sequence in self.sequences:
me@261	526 for chain in sequence.pdb_chains:
me@261	527 yield (sequence, chain)
me@270	528
me@261	529 def superimpose(self):
me@261	530 """ Superimpose all pdb_chains in this block """
me@261	531 sequences_chains = list(self.sequences_chains())
me@261	532 if len(sequences_chains) >= 1:
me@261	533 sup = Superimposer()
me@261	534 fixed_sequence, fixed_chain = sequences_chains.pop()
me@261	535 fixed_atoms = self.ca_atoms(fixed_sequence, fixed_chain)
me@261	536 for sequence, chain in sequences_chains:
me@261	537 moving_atoms = self.ca_atoms(sequence, chain)
me@261	538 sup.set_atoms(fixed_atoms, moving_atoms)
me@261	539 # Apply rotation/translation to the moving atoms
me@261	540 sup.apply(moving_atoms)
me@270	541
me@261	542 def pdb_save(self, out_file):
me@270	543 """ Save all sequences
me@270	544
me@261	545 Returns {(sequence, chain): CHAIN}
me@261	546 CHAIN is chain letter in new file
me@261	547 """
me@261	548 tmp_file = NamedTemporaryFile(delete=False)
me@261	549 tmp_file.close()
me@270	550
me@261	551 for sequence, chain in self.sequences_chains():
me@261	552 sequence.pdb_save(tmp_file.name, chain)
me@261	553 # TODO: read from tmp_file.name
me@261	554 # change CHAIN
me@261	555 # add to out_file
me@270	556
me@261	557 os.unlink(NamedTemporaryFile)
bnagaev@239	558
me@260	559 # vim: set ts=4 sts=4 sw=4 et: