00001 """ Contains classes to deal with generic sequence alignment stuff not specific to a particular program or format. classes: o Alignment """ # standard library import string # biopython from Bio.Seq import Seq from Bio.SeqRecord import SeqRecord from Bio import Alphabet from Bio.Alphabet import IUPAC 00018 class Alignment: """Represent a set of alignments. This is a base class to represent alignments, which can be subclassed to deal with an alignment in a specific format. """ 00024 def __init__(self, alphabet): """Initialize a new Alignment object. Arguments: o alphabet - The alphabet to use for the sequence objects that are created. This alphabet must be a gapped type. """ self._alphabet = alphabet # hold everything at a list of seq record objects self._records =  00035 def _str_line(self, record) : """Returns a truncated string representation of a SeqRecord. This is a PRIVATE function used by the __str__ method. """ if len(record.seq) <= 50 : return "%s %s" % (record.seq, record.id) else : return "%s...%s %s" \ % (record.seq[:44], record.seq[-3:], record.id) 00046 def __str__(self) : """Returns a multi-line string summary of the alignment. This output is intended to be readable, but large alignments are shown truncated. A maximum of 20 rows (sequences) and 50 columns are shown, with the record identifiers. This should fit nicely on a single screen. e.g. DNAAlphabet() alignment with 3 rows and 14 columns ACGATCAGCTAGCT Alpha CCGATCAGCTAGCT Beta ACGATGAGCTAGCT Gamma """ rows = len(self._records) lines = ["%s alignment with %i rows and %i columns" \ % (str(self._alphabet), rows, self.get_alignment_length())] if rows <= 20 : lines.extend([self._str_line(rec) for rec in self._records]) else : lines.extend([self._str_line(rec) for rec in self._records[:18]]) lines.append("...") lines.append(self._str_line(self._records[-1])) return "\n".join(lines) 00070 def __repr__(self) : """Returns a representation of the object for debugging. The representation cannot be used with eval() to recreate the object, which is usually possible with simple python ojects. For example: <Bio.Align.Generic.Alignment instance (2 records of length 14, SingleLetterAlphabet()) at a3c184c> The hex string is the memory address of the object, see help(id). This provides a simple way to visually distinguish alignments of the same size. """ return "<%s instance (%i records of length %i, %s) at %x>" % \ (self.__class__, len(self._records), self.get_alignment_length(), repr(self._alphabet), id(self)) #This version is useful for doing eval(repr(alignment)), #but it can be VERY long: #return "%s(%s, %s)" \ # % (self.__class__, repr(self._records), repr(self._alphabet)) 00091 def get_all_seqs(self): """Return all of the sequences involved in the alignment. The return value is a list of SeqRecord objects. """ return self._records 00098 def __iter__(self) : """Iterate over alignment rows as SeqRecord objects e.g. for record in align : print record.id print record.seq """ return iter(self._records) 00109 def get_seq_by_num(self, number): """Retrieve a sequence by row number. Returns: o A Seq object for the requested sequence. Raises: o IndexError - If the specified number is out of range. """ return self._records[number].seq 00120 def get_alignment_length(self): """Return the maximum length of the alignment. All objects in the alignment should (hopefully) have the same length. This function will go through and find this length by finding the maximum length of sequences in the alignment. """ max_length = 0 for record in self._records: if len(record.seq) > max_length: max_length = len(record.seq) return max_length 00135 def add_sequence(self, descriptor, sequence, start = None, end = None, weight = 1.0): """Add a sequence to the alignment. This doesn't do any kind of alignment, it just adds in the sequence object, which is assumed to be prealigned with the existing sequences. Arguments: o descriptor - The descriptive id of the sequence being added. This will be used as the resulting SeqRecord's .id property (and, for historical compatibility, also the .description property) o sequence - A string with sequence info. o start - You can explicitly set the start point of the sequence. This is useful (at least) for BLAST alignments, which can just be partial alignments of sequences. o end - Specify the end of the sequence, which is important for the same reason as the start. o weight - The weight to place on the sequence in the alignment. By default, all sequences have the same weight. (0.0 => no weight, 1.0 => highest weight) """ new_seq = Seq(sequence, self._alphabet) #We are now effectively using the SeqRecord's .id as #the primary identifier (e.g. in Bio.SeqIO) so we should #populate it with the descriptor. #For backwards compatibility, also store this in the #SeqRecord's description property. new_record = SeqRecord(new_seq, id = descriptor, description = descriptor) # hack! We really need to work out how to deal with annotations # and features in biopython. Right now, I'll just use the # generic annotations dictionary we've got to store the start # and end, but we should think up something better. I don't know # if I'm really a big fan of the LocatableSeq thing they've got # in BioPerl, but I'm not positive what the best thing to do on # this is... if start: new_record.annotations['start'] = start if end: new_record.annotations['end'] = end # another hack to add weight information to the sequence new_record.annotations['weight'] = weight self._records.append(new_record) 00186 def get_column(self,col): """Returns a string containing a given column.""" col_str = '' assert col >= 0 and col <= self.get_alignment_length() for rec in self._records: col_str += rec.seq[col] return col_str if __name__ == "__main__" : print "Mini self test..." raw_data = ["ACGATCAGCTAGCT", "CCGATCAGCTAGCT", "ACGATGAGCTAGCT"] a = Alignment(Alphabet.generic_dna) a.add_sequence("Alpha", raw_data, weight=2) a.add_sequence("Beta", raw_data) a.add_sequence("Gamma", raw_data) print print "str(a):" print str(a) print print "repr(a):" print repr(a) print #Iterating over the rows... for rec in a : assert isinstance(rec, SeqRecord) for r,rec in enumerate(a) : assert isinstance(rec, SeqRecord) assert raw_data[r] == rec.seq.tostring() if r==0 : assert rec.annotations['weight']==2 print "Alignment iteraction as SeqRecord OK"