def test_emits_one_single_color_unitig(self):
# given
b = CortexGraphBuilder()
b.with_colors(0)
b.add_edge('AAA', 'AAT', color=0)
b.make_consistent('AAA')
graph = b.build()
# when
paths = list(Interactor(graph).all_simple_paths())
# then
assert ['AAAT'] == [str(p.seq) for p in paths]
def test_two_colors_to_one_color_is_node_end(self, start_node, orientation):
# given
b = CortexGraphBuilder()
b.with_colors(0, 1)
b.with_node_coverage('AAA', 1, 1)
b.with_node_coverage('AAT', 1, 0)
b.add_edge_with_color('AAA', 'AAT', 0)
# when
graph = b.build()
# then
search = UnitigSearch.from_node_and_graph(start_node, graph)
assert search.is_unitig_end(start_node, orientation)
def test_follows_two_colors_with_no_color_specified(self):
# given
b = CortexGraphBuilder()
b.with_colors(0, 1)
b.add_edge('AAA', 'AAT', color=0)
b.add_edge('AAT', 'ATA', color=1)
b.make_consistent('AAA')
graph = b.build()
# when
paths = list(Interactor(graph).all_simple_paths())
# then
assert {'AAATA'} == set([str(p.seq) for p in paths])
def test_with_link_for_y_graph_emits_one_path(self):
# given
b = CortexGraphBuilder()
b.with_kmer_size(3)
b.add_path('AAA', 'AAC')
b.add_path('AAA', 'AAT')
b.make_consistent('AAA')
cdb = b.build()
links = LinksBuilder() \
.with_link_for_kmer('F 1 1 C', 'AAA') \
.build()
# when
paths = list(Interactor(cdb).all_simple_paths(links=links))
# then
assert ['AAAC'] == [str(p.seq) for p in paths]
def test_in_y_graph_finds_two_paths(self):
# given
b = CortexGraphBuilder()
b.add_path('CAA', 'AAA')
b.add_path('TAA', 'AAA')
b.make_consistent('AAA')
cdb = b.build()
# when
paths = list(Interactor(cdb).all_simple_paths())
# then
assert {'CAAA', 'TAAA'} == set([str(p.seq) for p in paths])
def test_two_connected_nodes_are_a_unitig(self, num_colors):
# given
colors = list(range(num_colors))
coverage = [1] + [0 for _ in range(1, num_colors)]
b = CortexGraphBuilder()
b.with_colors(*colors)
b.with_node_coverage('AAA', *coverage)
b.with_node_coverage('AAT', *coverage)
b.add_edge_with_color('AAA', 'AAT', 0)
# when
graph = b.build()
# then
search = UnitigSearch.from_node_and_graph('AAA', graph)
assert not search.is_unitig_end('AAA', EdgeTraversalOrientation.original)
assert search.is_unitig_end('AAA', EdgeTraversalOrientation.reverse)
search = UnitigSearch.from_node_and_graph('AAT', graph)
assert search.is_unitig_end('AAT', EdgeTraversalOrientation.original)
assert not search.is_unitig_end('AAT', EdgeTraversalOrientation.reverse)
def test_returns_two_nodes_in_order_regardless_of_starting_node(self, start_node):
# given
b = CortexGraphBuilder()
b.with_colors(0)
b.add_path('AAA', 'AAT', 'ATA')
# when
graph = b.build()
# then
search = UnitigSearch.from_node_and_graph(start_node, graph)
assert ['AAT', 'ATA'] == list(
search.next_unitig_node_iter(EdgeTraversalOrientation.original,
start_node='AAA'))
assert ['AAT', 'AAA'] == list(search.next_unitig_node_iter(EdgeTraversalOrientation.reverse,
start_node='ATA'))
def test_multi_color_y_graph_always_has_three_unitigs(self, data, num_colors, orientation):
# given
colors = list(range(num_colors))
edge_colors = [data.draw(s.sampled_from(colors)) for _ in range(3)]
b = CortexGraphBuilder()
b.with_colors(*colors)
for edge_color in edge_colors:
b.add_edge_with_color_and_coverage('AAA', 'AAT', edge_color, color_coverage=1)
b.add_edge_with_color_and_coverage('AAT', 'ATA', edge_colors[1], color_coverage=1)
b.add_edge_with_color_and_coverage('AAT', 'ATC', edge_colors[2], color_coverage=1)
# when
graph = b.build()
# then
for node in ['ATA', 'ATC']:
search = UnitigSearch.from_node_and_graph(node, graph)
search.is_unitig_end(node, orientation)
search = UnitigSearch.from_node_and_graph('AAA', graph)
assert not search.is_unitig_end('AAA', EdgeTraversalOrientation.original)
assert search.is_unitig_end('AAA', EdgeTraversalOrientation.reverse)
assert search.is_unitig_end('AAT', EdgeTraversalOrientation.original)
assert not search.is_unitig_end('AAT', EdgeTraversalOrientation.reverse)
def test_prunes_three_tips_of_length_1_reverse_kmer(self, seed_and_expected_kmer):
# given
seed, expected_kmer = seed_and_expected_kmer
b = CortexGraphBuilder()
b.with_colors(0, 1)
kmers = ['AAC', 'ACT', 'AAG', 'CAG']
for kmer in kmers:
b.add_node(kmer)
b.add_edge(kmers[0], kmers[1], 0)
b.add_edge(kmers[0], kmers[1], 1)
b.add_edge(kmers[1], kmers[2], 0)
b.add_edge(kmers[1], kmers[3], 1)
graph = b.build()
# when
graph = interactor.Interactor(graph).prune_tips_less_than(2).graph
# then
assert 1 == len(graph)
assert {expected_kmer} == set(graph.nodes)
def test_prunes_one_tip_of_length_1_in_y(self):
# given
b = CortexGraphBuilder()
b.with_colors(0, 1)
kmers = ['AAA', 'AAC', 'ACC', 'CCC', 'ACG', 'CGC']
for kmer in kmers:
b.add_node(kmer)
b.add_edge('AAA', 'AAC', 0)
b.add_edge('AAA', 'AAC', 1)
b.add_edge('AAC', 'ACC', 0)
b.add_edge('ACC', 'CCC', 0)
b.add_edge('AAC', 'ACG', 1)
b.add_edge('ACG', 'CGC', 1)
graph = b.build()
# when
graph = interactor.Interactor(graph) \
.prune_tips_less_than(2) \
.graph
# then
assert set(kmers[1:]) == set(graph.nodes)
assert 5 == len(graph)
def test_prunes_three_tips_of_length_1(self):
# given
b = CortexGraphBuilder()
b.with_colors(0, 1)
kmers = ['AAA', 'AAC', 'ACC', 'ACG']
for kmer in kmers:
b.add_node(kmer)
b.add_edge(kmers[0], kmers[1], 0)
b.add_edge(kmers[0], kmers[1], 1)
b.add_edge(kmers[1], kmers[2], 0)
b.add_edge(kmers[1], kmers[3], 1)
graph = b.build()
# when
graph = interactor.Interactor(graph) \
.prune_tips_less_than(2) \
.graph
# then
assert 1 == len(graph)
assert {'AAC'} == set(graph.nodes)
def test_two_colors_to_two_nodes_in_first_color_is_node_end(self):
# given
b = CortexGraphBuilder()
b.with_colors(0, 1)
b.with_node_coverage('AAA', 1, 1)
b.with_node_coverage('AAT', 1, 0)
b.with_node_coverage('AAC', 1, 0)
b.add_edge_with_color('AAA', 'AAT', 0)
b.add_edge_with_color('AAA', 'AAC', 0)
# when
graph = b.build()
# then
for orientation in EdgeTraversalOrientation:
for node in ['AAA', 'AAT', 'AAC']:
search = UnitigSearch.from_node_and_graph(node, graph)
assert search.is_unitig_end(node, orientation)
def test_middle_unconnected_node_for_one_color_in_two_color_graph_results_in_three_unitigs(
self, start_node, orientation
):
# given
b = CortexGraphBuilder()
b.with_colors(0, 1)
b.with_node_coverage('AAA', 1, 1)
b.with_node_coverage('AAT', 1, 1)
b.with_node_coverage('ATA', 1, 1)
b.add_edge_with_color('AAA', 'AAT', 0)
b.add_edge_with_color('AAT', 'ATA', 0)
# when
graph = b.build()
search = UnitigSearch.from_node_and_graph(start_node, graph)
assert search.is_unitig_end(start_node, orientation)
def test_bubble_and_y_with_two_links_returns_three_transcripts(self):
# given
links = LinksBuilder() \
.with_link_for_kmer('F 2 1 CT', 'AAA') \
.with_link_for_kmer('F 1 1 G', 'AAA') \
.build()
b = CortexGraphBuilder()
b.with_kmer_size(3)
b.add_path('AAA', 'AAC', 'ACC', 'CCC', 'CCA')
b.add_path('AAA', 'AAG', 'AGC', 'GCC', 'CCC', 'CCT')
b.make_consistent('AAA')
cdb = b.build()
# when
paths = list(Interactor(cdb).all_simple_paths(links=links))
# then
assert ['AAACCCT', 'AAAGCCCA',
'AAAGCCCT'] == sorted([str(p.seq) for p in paths])