def genomic_upset(options, label_names):
'''
Arguments:
input_files - List of BED files to to calculate the weights
output - output path
label_names - names of input files
Takes a list of sets a list of the sizes of non-overlapping intersections between them
'''
input_files = options.input
output = options.output
kwargs = hlp.map_bedtools_options(options.bedtools_options)
N = len(input_files)
# Generate a truth table of intersections to calculate
truth_table = [x for x in itertools.product("01", repeat=N)][1:]
weights = {}
for t in truth_table:
ones = [BedTool(input_files[i]) for i in range(N) if t[i] == '1']
zeros = [BedTool(input_files[i]) for i in range(N) if t[i] == '0']
#report those entries in set A which do ovelap with other sets
x = ones[0]
if len(ones) > 1:
for bed in ones[1:]:
x = x.intersect(bed, u=True, **kwargs)
#report those entries in set A which doesn't ovelap with other sets
if len(zeros) > 0:
#y = zeros[0]
for bed in zeros[0:]:
x = x.intersect(bed, v=True, **kwargs)
X = (x).count()
weights[''.join(t)] = X
#save the intersected results
if options.saveoverlaps:
if X >= options.overlapthresh:
file_name = ''
name_itr = 0
for name in t:
if name == '1':
file_name += '_' + label_names[name_itr]
name_itr += 1
file_name = ''.join(t) + file_name
hlp.create_dir(output + '/sets')
x.moveto(output + '/sets/' + file_name + '.bed')
#delete all temp files
helpers.cleanup()
return (weights)
def list_upset(options, label_names):
'''
Arguments:
input_files - List of list files to calculate weights for upset plot
output - output path
label_names - names of input files
Takes a list of sets a list of the sizes of non-overlapping intersections between them
'''
input_files = options.input
output = options.output
S = []
for f in input_files:
with open(f) as f_open:
S.append(set(f_open.read().splitlines()))
N = len(S)
# Generate a truth table of intersections to calculate
truth_table = [x for x in itertools.product("01", repeat=N)][1:]
weights = {}
for t in truth_table:
ones = [S[i] for i in range(N) if t[i] == '1']
zeros = [S[i] for i in range(N) if t[i] == '0']
X = set.intersection(*ones)
X.difference_update(*zeros)
weights[''.join(t)] = len(X)
#save the intersected results
if options.saveoverlaps:
if len(X) >= options.overlapthresh:
file_name = ''
name_itr = 0
for name in t:
if name == '1':
file_name += '_' + label_names[name_itr]
name_itr += 1
file_name = ''.join(t) + file_name
hlp.create_dir(output + '/sets')
inter_file = open(output + '/sets/' + file_name + '.txt', 'w')
inter_file.writelines('\n'.join(list(X)))
inter_file.close()
return (weights)