def get_other_consensi_all(pname, fragment, VERBOSE=0):
'''Get all other consensi except the ones from this patient'''
from Bio import SeqIO
from hivwholeseq.patients.filenames import get_consensi_alignment_filename
fn = get_consensi_alignment_filename('all', fragment)
consensi = []
for seq in SeqIO.parse(fn, 'fasta'):
# NOTE: the first of seq.name.split('_') is the patient name
if seq.name.split('_')[0] != pname:
seq.seq = seq.seq.ungap('-')
consensi.append(seq)
return consensi
def get_other_consensi_all(pname, fragment, VERBOSE=0):
'''Get all other consensi except the ones from this patient'''
from Bio import SeqIO
from hivwholeseq.patients.filenames import get_consensi_alignment_filename
fn = get_consensi_alignment_filename('all', fragment)
consensi = []
for seq in SeqIO.parse(fn, 'fasta'):
# NOTE: the first of seq.name.split('_') is the patient name
if seq.name.split('_')[0] != pname:
seq.seq = seq.seq.ungap('-')
consensi.append(seq)
return consensi
def get_consensi_alignment_filename(self, region, format='fasta'):
'''Get the filename of the multiple sequence alignment of all consensi'''
from hivwholeseq.patients.filenames import get_consensi_alignment_filename
return get_consensi_alignment_filename(self.name,
region,
format=format)
samples = load_samples_sequenced()
if seq_runs is not None:
samples = samples.loc[samples['seq run'].isin(seq_runs)]
if adaIDs is not None:
samples = samples.loc[samples.adapter.isin(adaIDs)]
if use_pats:
samples = samples.loc[samples['patient sample'] != 'nan']
else:
samples = samples.loc[samplenames]
if fragments is None:
fragments = ['F'+str(i+1) for i in xrange(6)]
alis = {fr: AlignIO.read(get_consensi_alignment_filename('all', fr), 'fasta')
for fr in fragments}
for samplename, sample in samples.iterrows():
sample = SampleSeq(sample)
data_folder = sample.seqrun_folder
adaID = sample.adapter
pname = sample.patientname
for fragment in fragments:
if VERBOSE >= 1:
print sample['seq run'], adaID, fragment, samplename,
# Read the summary filename of the filter_mapped, and find out whether
# there are many distant reads (a few are normal)
fn = get_filter_mapped_summary_filename(data_folder, adaID, fragment)
'subtype', 'confidence'),
)
write_json(tree_json, fn, indent=1)
if use_joint:
if VERBOSE >= 2:
print 'Align all patients',
ali_all = align_muscle(*seqs_all, sort=True)
if VERBOSE >= 2:
print 'OK'
if use_save:
if VERBOSE >= 2:
print 'Save all patients',
reg_tmp = '_'.join(pcodes) + '_' + region
fn_out = get_consensi_alignment_filename('all', reg_tmp)
mkdirs(os.path.dirname(fn_out))
AlignIO.write(ali, fn_out, 'fasta')
if VERBOSE >= 2:
print 'OK'
if VERBOSE >= 2:
print 'Build local tree'
tree = build_tree_fasttree(ali_all, VERBOSE=VERBOSE)
if VERBOSE >= 2:
print 'Infer ancestral sequences'
a = ancestral_sequences(tree,
ali_all,
alphabet='ACGT-N',
copy_tree=False,
)
write_json(tree_json, fn, indent=1)
if use_joint:
if VERBOSE >= 2:
print 'Align all patients',
ali_all = align_muscle(*seqs_all, sort=True)
if VERBOSE >= 2:
print 'OK'
if use_save:
if VERBOSE >= 2:
print 'Save all patients',
reg_tmp = '_'.join(pcodes)+'_'+region
fn_out = get_consensi_alignment_filename('all', reg_tmp)
mkdirs(os.path.dirname(fn_out))
AlignIO.write(ali, fn_out, 'fasta')
if VERBOSE >= 2:
print 'OK'
if VERBOSE >= 2:
print 'Build local tree'
tree = build_tree_fasttree(ali_all, VERBOSE=VERBOSE)
if VERBOSE >= 2:
print 'Infer ancestral sequences'
a = ancestral_sequences(tree, ali_all, alphabet='ACGT-N', copy_tree=False,
attrname='sequence', seqtype='str')
a.calc_ancestral_sequences()
a.cleanup_tree()
def get_consensi_alignment_filename(self, region, format='fasta'):
'''Get the filename of the multiple sequence alignment of all consensi'''
from hivwholeseq.patients.filenames import get_consensi_alignment_filename
return get_consensi_alignment_filename(self.name, region, format=format)
