def test_filter_tiling_paths_by_len():
p_ctg_tiling_path_file = os.path.join(
helpers.get_test_data_dir(), 'p_ctg_tiling_path_1')
p_path, p_edge_to_ctg = mod.load_tiling_paths(p_ctg_tiling_path_file, 'P')
_, p_ctg_len = mod.calc_tiling_paths_len(p_path)
p_path_filtered = mod.filter_tiling_paths_by_len(p_path, p_ctg_len, 0)
assert(sorted(p_path_filtered.keys()) == sorted(
['000000F', '000001F', '000002F', '000003F', '000004F']))
p_path_filtered = mod.filter_tiling_paths_by_len(p_path, p_ctg_len, 10000)
assert(sorted(p_path_filtered.keys()) == sorted(
['000000F', '000001F', '000002F', '000003F']))
p_path_filtered = mod.filter_tiling_paths_by_len(p_path, p_ctg_len, 35000)
assert(sorted(p_path_filtered.keys()) == sorted(['000000F', '000001F']))
p_path_filtered = mod.filter_tiling_paths_by_len(p_path, p_ctg_len, 100000)
assert(sorted(p_path_filtered.keys()) == sorted([]))
# Test a degenerate case where there is no length for a particular contig.
keys = p_ctg_len.keys()
p_ctg_len_degenerate = {}
for i in xrange(1, len(keys)):
p_ctg_len_degenerate[keys[i]] = p_ctg_len[keys[i]]
with pytest.raises(Exception) as e_info:
p_path_filtered = mod.filter_tiling_paths_by_len(
p_path, p_ctg_len_degenerate, 0)
def test_format_gfa_v1_path_line():
gfa_graph = mod.GFAGraph()
# Load tiling paths from file.
p_ctg_tiling_path_file = os.path.join(helpers.get_test_data_dir(),
'p_ctg_tiling_path_1')
p_paths, p_edge_to_ctg = gen_gfa_v1.load_tiling_paths(
p_ctg_tiling_path_file, 'P')
# If seq_len_map is None, all CIGAR operations should be '*'.
expected = {
'000000F':
'P\t000000F\t000092122-,000081654-,000034462-,000061403-,000021348-,000062240-,000083779-,000019819+,000063672+,000026565+,000050047-\t*,*,*,*,*,*,*,*,*,*,*',
'000001F':
'P\t000001F\t000070651+,000018109+,000068978+,000100559+,000010548-,000006846-,000065052-,000071922+,000076878+,000000861+,000001755-\t*,*,*,*,*,*,*,*,*,*,*',
'000002F':
'P\t000002F\t000088930+,000008918+,000100248-,000085315-,000071965+,000082497+\t*,*,*,*,*,*',
'000003F': 'P\t000003F\t000084518+,000011674+,000057445-\t*,*,*',
'000004F': 'P\t000004F\t000014727+,000024020+,000060868+\t*,*,*',
}
seq_len_map = None
for ctg_id, path in p_paths.iteritems():
path_line = gfa_graph.format_gfa_v1_path_line(ctg_id, path,
seq_len_map)
assert (path_line == expected[ctg_id])
# The seq_len_map dict is only used for the first read in the path,
# because it needs to be included completely. The other CIGAR operations
# are determined directly from the edges.
expected = {
'000000F':
'P\t000000F\t000092122-,000081654-,000034462-,000061403-,000021348-,000062240-,000083779-,000019819+,000063672+,000026565+,000050047-\t10000M,33726M,10123M,1352M,9924M,5834M,862M,5562M,1384M,473M,2171M',
'000001F':
'P\t000001F\t000070651+,000018109+,000068978+,000100559+,000010548-,000006846-,000065052-,000071922+,000076878+,000000861+,000001755-\t10000M,10077M,3766M,2648M,2421M,2089M,18168M,2723M,2451M,666M,15088M',
'000002F':
'P\t000002F\t000088930+,000008918+,000100248-,000085315-,000071965+,000082497+\t10000M,15215M,3113M,4851M,1857M,6035M',
'000003F':
'P\t000003F\t000084518+,000011674+,000057445-\t10000M,9432M,23096M',
'000004F':
'P\t000004F\t000014727+,000024020+,000060868+\t10000M,5238M,3235M',
}
for ctg_id, path in p_paths.iteritems():
# Initialize all reads to a fixed value, just to be safe.
seq_len_map = {}
for edge in path:
v, w = edge[0], edge[1]
seq_len_map[v.split(':')[0]] = 10000
seq_len_map[w.split(':')[0]] = 10000
path_line = gfa_graph.format_gfa_v1_path_line(ctg_id, path,
seq_len_map)
assert (path_line == expected[ctg_id])
# Test a degenerate case where path is None.
path_line = gfa_graph.format_gfa_v1_path_line('', None, None)
assert (path_line == '')
def test_calc_tiling_paths_len():
p_ctg_tiling_path_file = os.path.join(
helpers.get_test_data_dir(), 'p_ctg_tiling_path_1')
p_path, p_edge_to_ctg = mod.load_tiling_paths(p_ctg_tiling_path_file, 'P')
p_coords, p_ctg_len = mod.calc_tiling_paths_len(p_path)
for ctg_id in p_coords.keys():
shared_items = set(expected_coord_map[ctg_id].items()) & set(
p_coords[ctg_id].items())
assert(len(shared_items) == len(p_coords[ctg_id]))
assert(expected_contig_len[ctg_id] == p_ctg_len[ctg_id])
def test_calc_node_coords():
# The p_ctg_tiling_path_1 is a normal tiling path file.
p_ctg_tiling_path_file = os.path.join(
helpers.get_test_data_dir(), 'p_ctg_tiling_path_1')
p_paths, p_edge_to_ctg = mod.load_tiling_paths(p_ctg_tiling_path_file, 'P')
ctg_id = '000000F'
coord_map, contig_len = mod.calc_node_coords(p_paths[ctg_id])
shared_items = set(expected_coord_map[ctg_id].items()) & set(
coord_map.items())
assert(len(shared_items) == len(coord_map))
assert(expected_contig_len[ctg_id] == contig_len)
# The p_ctg_tiling_path_2 has two degenerative cases:
# - 000000F which has an inner cycle
# - 000001F which has an out-of-order edge
# - 000002F which is circular (this is a valid case)
p_ctg_tiling_path_file = os.path.join(
helpers.get_test_data_dir(), 'p_ctg_tiling_path_2')
p_paths, p_edge_to_ctg = mod.load_tiling_paths(p_ctg_tiling_path_file, 'P')
# Allow cycles, but the node's coord gets overwritten.
ctg_id = '000000F'
coord_map, contig_len = mod.calc_node_coords(p_paths[ctg_id])
assert(coord_map['000081654:B'] == 55125)
# Do not allow unsorted graphs.
ctg_id = '000001F'
with pytest.raises(Exception) as e_info:
coord_map, contig_len = mod.calc_node_coords(p_paths[ctg_id])
# Allow circular graphs.
ctg_id = '000002F'
coord_map, contig_len = mod.calc_node_coords(p_paths[ctg_id])
assert(contig_len == 18473)
# Test for an empty tiling path.
coord_map, contig_len = mod.calc_node_coords([])
assert(not coord_map)
assert(contig_len == 0)
def test_load_tiling_paths():
p_ctg_tiling_path_file = os.path.join(
helpers.get_test_data_dir(), 'p_ctg_tiling_path_1')
p_path, p_edge_to_ctg = mod.load_tiling_paths(p_ctg_tiling_path_file, 'P')
assert(sorted(p_path.keys()) == sorted(
['000000F', '000001F', '000002F', '000003F', '000004F']))
for ctg_id, path in p_path.iteritems():
for edge in path:
v, w, b, e, l, idt, etype = edge
assert((v, w) in p_edge_to_ctg)
assert(p_edge_to_ctg[(v, w)] == (ctg_id, etype))
def test_add_tiling_path():
# Load the tiling path. These methods are tested in test_gen_gfa_v1.py.
p_ctg_tiling_path_file = os.path.join(
helpers.get_test_data_dir(), 'p_ctg_tiling_path_1')
p_paths, p_edge_to_ctg = gen_gfa_v1.load_tiling_paths(
p_ctg_tiling_path_file, 'P')
# Create a new GFA graph.
gfa_graph = mod.GFAGraph()
# Add the tiling paths.
for ctg_id, path in p_paths.iteritems():
gfa_graph.add_tiling_path(path, ctg_id)
# Check if we have the correct number of tiling paths.
assert(len(gfa_graph.paths.keys()) == len(p_paths.keys()))
# They should be same as loaded.
for ctg_id, path in p_paths.iteritems():
assert(ctg_id in gfa_graph.paths)
assert(gfa_graph.paths[ctg_id] == path)
def test_write_gfa_v1_2():
# Tests a case where a node is added to the graph, but
# there is no corresponding pread in preads4falcon.fasta file.
# Create a GFA graph.
gfa_graph = mod.GFAGraph()
# Load the p_ctg tiling paths.
p_ctg_tiling_path_file = os.path.join(
helpers.get_test_data_dir(), 'gfa-1', 'p_ctg_tiling_path')
p_paths, p_edge_to_ctg = gen_gfa_v1.load_tiling_paths(
p_ctg_tiling_path_file, 'P')
# Add the tiling paths to the GFA.
for ctg_id, path in p_paths.iteritems():
gfa_graph.add_tiling_path(path, ctg_id)
# Init paths to other input files.
preads_file = os.path.join(
helpers.get_test_data_dir(), 'gfa-1', 'preads4falcon.fasta')
p_ctg_fasta = os.path.join(
helpers.get_test_data_dir(), 'gfa-1', 'p_ctg.fa')
a_ctg_fasta = os.path.join(
helpers.get_test_data_dir(), 'gfa-1', 'a_ctg.fa')
write_reads = False
write_contigs = False
fp_out = StringIO()
# Add a node which does not exist in the preads4falcon.fasta file.
gfa_graph.add_read_from_node('12345:B')
# Run the unit under test.
with pytest.raises(Exception) as e_info:
gfa_graph.write_gfa_v1(fp_out, preads_file, [
p_ctg_fasta, a_ctg_fasta], write_reads, write_contigs)
def wrap_write_gfa_v1_test(use_sg, use_nx, use_tp, write_reads, write_contigs, min_p_len, min_a_len, expected_path):
# Create a GFA graph.
gfa_graph = mod.GFAGraph()
if use_sg:
# Load the assembly graph.
sg_edges_list = os.path.join(
helpers.get_test_data_dir(), 'gfa-1', 'sg_edges_list')
utg_data = os.path.join(
helpers.get_test_data_dir(), 'gfa-1', 'utg_data')
ctg_paths = os.path.join(
helpers.get_test_data_dir(), 'gfa-1', 'ctg_paths')
asm_graph = AsmGraph(sg_edges_list, utg_data, ctg_paths)
# Add the string graph to the GFA.
gfa_graph.add_asm_graph(asm_graph)
if use_tp:
# Load the p_ctg tiling paths.
p_ctg_tiling_path_file = os.path.join(
helpers.get_test_data_dir(), 'gfa-1', 'p_ctg_tiling_path')
p_paths, p_edge_to_ctg = gen_gfa_v1.load_tiling_paths(
p_ctg_tiling_path_file, 'P')
# Add the tiling paths to the GFA.
for ctg_id, path in p_paths.iteritems():
_, contig_len = gen_gfa_v1.calc_node_coords(path)
if contig_len >= min_p_len:
gfa_graph.add_tiling_path(path, ctg_id)
a_ctg_tiling_path_file = os.path.join(
helpers.get_test_data_dir(), 'gfa-1', 'a_ctg_tiling_path')
a_paths, a_edge_to_ctg = gen_gfa_v1.load_tiling_paths(
a_ctg_tiling_path_file, 'P')
# Add the tiling paths to the GFA.
for ctg_id, path in a_paths.iteritems():
_, contig_len = gen_gfa_v1.calc_node_coords(path)
if contig_len >= min_a_len:
gfa_graph.add_tiling_path(path, ctg_id)
if use_nx:
gexf_file = os.path.join(
helpers.get_test_data_dir(), 'gfa-1', 'sg.gexf')
nx_sg = nx.read_gexf(gexf_file)
gfa_graph.add_nx_string_graph(nx_sg)
# Init paths to other input files.
preads_file = os.path.join(
helpers.get_test_data_dir(), 'gfa-1', 'preads4falcon.fasta')
p_ctg_fasta = os.path.join(
helpers.get_test_data_dir(), 'gfa-1', 'p_ctg.fa')
a_ctg_fasta = os.path.join(
helpers.get_test_data_dir(), 'gfa-1', 'a_ctg.fa')
fp_out = StringIO()
# Run the unit under test.
gfa_graph.write_gfa_v1(fp_out, preads_file, [
p_ctg_fasta, a_ctg_fasta], write_reads, write_contigs)
# Compare results.
result = fp_out.getvalue()
result = result.splitlines()
expected = [line.strip() for line in open(expected_path).readlines()]
assert(result == expected)