Bio.AlignIO.NexusIO

29 """Returns SeqRecord objects from a Nexus file. 30 31 Thus uses the Bio.Nexus module to do the hard work. 32 33 You are expected to call this function via Bio.SeqIO or Bio.AlignIO 34 (and not use it directly). 35 36 NOTE - We only expect ONE alignment matrix per Nexus file, 37 meaning this iterator will only yield one MultipleSeqAlignment. 38 """ 39 n = Nexus.Nexus(handle) 40 if not n.matrix: 41 #No alignment found 42 raise StopIteration 43 44 #Bio.Nexus deals with duplicated names by adding a '.copy' suffix. 45 #The original names and the modified names are kept in these two lists: 46 assert len(n.unaltered_taxlabels) == len(n.taxlabels) 47 48 if seq_count and seq_count != len(n.unaltered_taxlabels): 49 raise ValueError("Found %i sequences, but seq_count=%i" 50 % (len(n.unaltered_taxlabels), seq_count)) 51 52 #ToDo - Can we extract any annotation too? 53 records = (SeqRecord(n.matrix[new_name], id=new_name, 54 name=old_name, description="") 55 for old_name, new_name 56 in zip(n.unaltered_taxlabels, n.taxlabels)) 57 #All done 58 yield MultipleSeqAlignment(records, n.alphabet)

62 """Nexus alignment writer. 63 64 Note that Nexus files are only expected to hold ONE alignment 65 matrix. 66 67 You are expected to call this class via the Bio.AlignIO.write() or 68 Bio.SeqIO.write() functions. 69 """

70 - def write_file(self, alignments):

71 """Use this to write an entire file containing the given alignments. 72 73 alignments - A list or iterator returning MultipleSeqAlignment objects. 74 This should hold ONE and only one alignment. 75 """ 76 align_iter = iter(alignments) # Could have been a list 77 try: 78 first_alignment = align_iter.next() 79 except StopIteration: 80 first_alignment = None 81 if first_alignment is None: 82 #Nothing to write! 83 return 0 84 85 #Check there is only one alignment... 86 try: 87 second_alignment = align_iter.next() 88 except StopIteration: 89 second_alignment = None 90 if second_alignment is not None: 91 raise ValueError("We can only write one Alignment to a Nexus file.") 92 93 #Good. Actually write the single alignment, 94 self.write_alignment(first_alignment) 95 return 1 # we only support writing one alignment!

96

97 - def write_alignment(self, alignment):

98 #Creates an empty Nexus object, adds the sequences, 99 #and then gets Nexus to prepare the output. 100 if len(alignment) == 0: 101 raise ValueError("Must have at least one sequence") 102 if alignment.get_alignment_length() == 0: 103 raise ValueError("Non-empty sequences are required") 104 minimal_record = "#NEXUS\nbegin data; dimensions ntax=0 nchar=0; " \ 105 + "format datatype=%s; end;" \ 106 % self._classify_alphabet_for_nexus(alignment._alphabet) 107 n = Nexus.Nexus(minimal_record) 108 n.alphabet = alignment._alphabet 109 for record in alignment: 110 n.add_sequence(record.id, str(record.seq)) 111 n.write_nexus_data(self.handle)

112

113 - def _classify_alphabet_for_nexus(self, alphabet):

114 """Returns 'protein', 'dna', 'rna' based on the alphabet (PRIVATE). 115 116 Raises an exception if this is not possible.""" 117 #Get the base alphabet (underneath any Gapped or StopCodon encoding) 118 a = Alphabet._get_base_alphabet(alphabet) 119 120 if not isinstance(a, Alphabet.Alphabet): 121 raise TypeError("Invalid alphabet") 122 elif isinstance(a, Alphabet.ProteinAlphabet): 123 return "protein" 124 elif isinstance(a, Alphabet.DNAAlphabet): 125 return "dna" 126 elif isinstance(a, Alphabet.RNAAlphabet): 127 return "rna" 128 else: 129 #Must be something like NucleotideAlphabet or 130 #just the generic Alphabet (default for fasta files) 131 raise ValueError("Need a DNA, RNA or Protein alphabet")

Source Code for Module Bio.AlignIO.NexusIO