class FisherRunner(object):
def __init__(self):
self.data_class = JointData
self.classes = ('Reference', 'Germline', 'Somatic', 'LOH', 'Unknown')
def run(self, args):
self.reader = JointCountsReader(args.jcnt_file_name)
self.writer = csv.writer(open(args.tsv_file_name, 'w'), delimiter='\t')
chr_list = self.reader.get_chr_list()
for chr_name in sorted(chr_list):
self._classify_chromosome(chr_name)
self.reader.close()
def _classify_chromosome(self, chr_name):
counts = self.reader.get_counts(chr_name)
jcnt_rows = self.reader.get_rows(chr_name)
end = self.reader.get_chr_size(chr_name)
n = int(1e5)
start = 0
stop = min(n, end)
while start < end:
sub_counts = counts[start:stop]
sub_rows = jcnt_rows[start:stop]
data = self.data_class(sub_counts)
labels = self.model.classify(data)
self._write_rows(chr_name, sub_rows, labels)
start = stop
stop = min(stop + n, end)
def _write_rows(self, chr_name, rows, labels):
for i, row in enumerate(rows):
out_row = [chr_name]
out_row.extend(row)
label = int(labels[i])
class_name = self.classes[label]
out_row.append(class_name)
if class_name == 'Somatic':
print out_row
self.writer.writerow(out_row)
def jcnt_to_cncnt( args ):
reader = JointCountsReader( args.jcnt_file_name )
chr_list = reader.get_chr_list()
cncnt_file = ConanCountsFile( args.cncnt_file_name, 'w' )
segment_reader = csv.reader( open( args.segment_file_name ), delimiter='\t' )
for row in segment_reader:
print row
chr_name = row[0]
if chr_name == '23':
chr_name = 'X'
if chr_name == '24':
chr_name = 'Y'
start = int( row[1] )
stop = int( row[2] )
cn_status = row[3]
if cn_status == '7':
cn_status = '1'
elif cn_status == '8':
cn_status = '2'
elif cn_status == '9':
cn_status = '4'
elif cn_status == '10':
cn_status = '5'
elif cn_status == '11':
cn_status = '6'
if chr_name not in chr_list:
continue
rows = reader.get_rows( chr_name )
segment_indices = np.logical_and( rows['position'] >= start, rows['position'] <= stop )
segment_rows = rows[segment_indices]
if len( segment_rows ) == 0:
continue
cncnt_file.add_rows( cn_status, chr_name, segment_rows )
reader.close()
cncnt_file.close()
