def retrieve_alignment(tre, alnpath, taxonset=range(0, 101), delimiter='_'):
"""
Parameters
----------------
tre : single-copy treeswift tree generated from James's code
alnpath : path to the phylip formatted alignment of the genes. The row labels should be a superset of the leafset of 'tre'
seqlen : sequence length parameter, only the first seqlen columns are taken from the MSA
taxonset: set, the taxon set of the entire dataset
Returns the MSA that corresponds to the input tree.
"""
aln = AlignIO.read(open(alnpath), "phylip")
seqlen = len(aln[0].seq)
blank = "-" * seqlen
whitelist = set(tre.labels(True, False))
rest = set(taxonset)
#print(rest)
res = MultipleSeqAlignment([])
for r in aln[:, :seqlen]:
if r.id in whitelist:
rid = r.id.split(delimiter)[0]
rid_i = rid
res.append(SeqRecord(r.seq, id=rid))
rest.remove(rid_i)
for rst in rest:
res.append(SeqRecord(Seq(blank), id=str(rst)))
res.sort()
return res
def dict_to_bioalignment(d, alphabet='generic_alphabet', sorted=True):
"""
Create a BioPython MultipleSequenceAlignment
from a dict with pairs consisting of id and sequence.
"""
alignment = MultipleSeqAlignment([])
bio_alphabet = getattr(Bio.Alphabet, alphabet)
for id, seq in d.items():
seq_record = SeqRecord(Seq(seq, alphabet=bio_alphabet), id=id)
alignment.append(seq_record)
if sorted:
alignment.sort()
return alignment
