def test_subpoagraph_should_omit_edges_2(self):
nodes = [
graph.Node(node_id=nid(0), base=b('A'), aligned_to=None),
graph.Node(node_id=nid(1), base=b('C'), aligned_to=None),
graph.Node(node_id=nid(2), base=b('C'), aligned_to=None)
]
sequences = {
msa.SequenceID('seq1'):
graph.Sequence(msa.SequenceID('seq1'),
[graph.SeqPath([*map(nid, [0, 2])])],
graph.SequenceMetadata({'group': '1'})),
msa.SequenceID('seq2'):
graph.Sequence(msa.SequenceID('seq2'),
[graph.SeqPath([*map(nid, [0, 1, 2])])],
graph.SequenceMetadata({'group': '1'}))
}
poagraph = graph.Poagraph(nodes, sequences)
translator = poa._PoagraphPOTranslator(poagraph,
[msa.SequenceID('seq1')])
actual_po_content = translator.get_input_po_content()
expected_po_content = "VERSION=pangenome\n" \
"NAME=pangenome\n" \
"TITLE=pangenome\n" \
"LENGTH=2\n" \
"SOURCECOUNT=1\n" \
"SOURCENAME=seq1\n" \
"SOURCEINFO=2 0 100 -1 seq1\n" \
"a:S0\n" \
"c:L0S0"
self.assertEqual(expected_po_content, actual_po_content)
def test_subpoagraph_construction_full_graph(self):
nodes = [
graph.Node(node_id=nid(0), base=b('A'), aligned_to=None),
graph.Node(node_id=nid(1), base=b('A'), aligned_to=None),
graph.Node(node_id=nid(2), base=b('C'), aligned_to=None),
graph.Node(node_id=nid(3), base=b('A'), aligned_to=None),
graph.Node(node_id=nid(4), base=b('T'), aligned_to=None)
]
sequences = {
msa.SequenceID('seq0'):
graph.Sequence(msa.SequenceID('seq0'),
[graph.SeqPath([*map(nid, [0, 1, 2, 3, 4])])],
graph.SequenceMetadata({'group': '1'}))
}
poagraph = graph.Poagraph(nodes, sequences)
translator = poa._PoagraphPOTranslator(poagraph,
[msa.SequenceID('seq0')])
actual_po_content = translator.get_input_po_content()
expected_po_content = "VERSION=pangenome\n" \
"NAME=pangenome\n" \
"TITLE=pangenome\n" \
"LENGTH=5\n" \
"SOURCECOUNT=1\n" \
"SOURCENAME=seq0\n" \
"SOURCEINFO=5 0 100 -1 seq0\n" \
"a:S0\n" \
"a:L0S0\n" \
"c:L1S0\n" \
"a:L2S0\n" \
"t:L3S0"
self.assertEqual(expected_po_content, actual_po_content)
def __init__(self):
self.sources = {
msa.SequenceID("seq0"): "",
msa.SequenceID("seq1"): "ACTAGGT",
msa.SequenceID("seq2"): "GGTCAGT",
msa.SequenceID("seq3"): "",
msa.SequenceID("seq4"): ""
}
def test_1_p_parameter_influence(self, p: at_params.P,
expected_cutoff: graph.Compatibility):
nodes = [
graph.Node(node_id=nid(0), base=b('T'), aligned_to=None),
graph.Node(node_id=nid(1), base=b('A'), aligned_to=None),
graph.Node(node_id=nid(2), base=b('G'), aligned_to=None),
graph.Node(node_id=nid(3), base=b('A'), aligned_to=None),
graph.Node(node_id=nid(4), base=b('C'), aligned_to=None),
graph.Node(node_id=nid(5), base=b('A'), aligned_to=None),
graph.Node(node_id=nid(6), base=b('C'), aligned_to=None),
graph.Node(node_id=nid(7), base=b('G'), aligned_to=None),
graph.Node(node_id=nid(8), base=b('T'), aligned_to=None),
graph.Node(node_id=nid(9), base=b('A'), aligned_to=None)
]
sequences = {
msa.SequenceID('seq0'):
graph.Sequence(msa.SequenceID('seq0'), [
graph.SeqPath(
[*map(nid, [10, 11, 12, 13, 14, 15, 16, 17, 18, 9])])
], graph.SequenceMetadata({})),
msa.SequenceID('seq1'):
graph.Sequence(msa.SequenceID('seq1'), [
graph.SeqPath(
[*map(nid, [10, 11, 12, 13, 14, 15, 16, 17, 8, 9])])
], graph.SequenceMetadata({})),
msa.SequenceID('seq2'):
graph.Sequence(msa.SequenceID('seq2'), [
graph.SeqPath(
[*map(nid, [10, 11, 12, 13, 14, 15, 16, 7, 8, 9])])
], graph.SequenceMetadata({})),
msa.SequenceID('seq3'):
graph.Sequence(msa.SequenceID('seq3'), [
graph.SeqPath([*map(nid, [10, 11, 12, 3, 4, 5, 6, 7, 8, 9])])
], graph.SequenceMetadata({})),
msa.SequenceID('seq4'):
graph.Sequence(
msa.SequenceID('seq3'),
[graph.SeqPath([*map(nid, [10, 11, 2, 3, 4, 5, 6, 7, 8, 9])])],
graph.SequenceMetadata({}))
}
poagraph = graph.Poagraph(nodes, sequences)
consensus_path = graph.SeqPath(
[*map(nid, [10, 11, 12, 13, 14, 15, 16, 17, 18, 19])])
compatibilities = poagraph.get_compatibilities(
poagraph.get_sequences_ids(), consensus_path, p)
actual_cutoff = at_builders._find_node_cutoff(
[c for c in compatibilities.values()], []).cutoff
self.assertAlmostEqual(expected_cutoff.value, actual_cutoff.value)
def test_04_single_block_no_nucleotides(self):
maf_path = self.maf_files_dir.joinpath(
"test_4_single_block_no_nucleotides.maf")
expected_nodes = []
expected_sequences = {
msa.SequenceID('seq0'):
graph.Sequence(msa.SequenceID('seq0'),
[],
graph.SequenceMetadata({'group': '1'})),
msa.SequenceID('seq1'):
graph.Sequence(msa.SequenceID('seq1'),
[],
graph.SequenceMetadata({'group': '1'})),
msa.SequenceID('seq2'):
graph.Sequence(msa.SequenceID('seq2'),
[],
graph.SequenceMetadata({'group': '2'})),
msa.SequenceID('seq3'):
graph.Sequence(msa.SequenceID('seq3'),
[],
graph.SequenceMetadata({'group': '2'}))
}
expected_poagraph = graph.Poagraph(expected_nodes, expected_sequences)
actual_poagraph, _ = builder.build_from_dagmaf(
msa.Maf(pathtools.get_file_content_stringio(maf_path), maf_path),
self.fasta_provider,
self.metadatacsv)
self.assertEqual(expected_poagraph, actual_poagraph)
def test_7_missing_one_reverted_sequence_middle_minus1_minus1(self):
maf_path = self.maf_files_dir.joinpath(
"test_7_missing_one_reverted_sequence_middle_minus1_minus1.maf")
expected_nodes = [
# block 0
graph.Node(node_id=nid(0), base=graph.Base('A'), aligned_to=None),
graph.Node(node_id=nid(1), base=graph.Base('C'), aligned_to=None),
graph.Node(node_id=nid(2), base=graph.Base('T'), aligned_to=None),
graph.Node(node_id=nid(3), base=graph.Base('A'), aligned_to=None),
# missing seq2
graph.Node(node_id=nid(4),
base=graph.Base(self.missing_n.value),
aligned_to=None),
graph.Node(node_id=nid(5),
base=graph.Base(self.missing_n.value),
aligned_to=None),
# block 1
graph.Node(node_id=nid(6), base=graph.Base('A'),
aligned_to=nid(7)),
graph.Node(node_id=nid(7), base=graph.Base('G'),
aligned_to=nid(6)),
graph.Node(node_id=nid(8), base=graph.Base('C'),
aligned_to=nid(9)),
graph.Node(node_id=nid(9), base=graph.Base('G'),
aligned_to=nid(8)),
graph.Node(node_id=nid(10),
base=graph.Base('C'),
aligned_to=nid(11)),
graph.Node(node_id=nid(11),
base=graph.Base('T'),
aligned_to=nid(10)),
]
expected_sequences = {
msa.SequenceID('seq0'):
graph.Sequence(msa.SequenceID('seq0'), [],
graph.SequenceMetadata({'group': '1'})),
msa.SequenceID('seq1'):
graph.Sequence(
msa.SequenceID('seq1'),
[graph.SeqPath([*map(nid, [0, 1, 2, 3, 7, 9, 11])])],
graph.SequenceMetadata({'group': '1'})),
msa.SequenceID('seq2'):
graph.Sequence(
msa.SequenceID('seq2'),
[graph.SeqPath([*map(nid, [0, 1, 4, 5, 6, 8, 10])])],
graph.SequenceMetadata({'group': '2'})),
msa.SequenceID('seq3'):
graph.Sequence(msa.SequenceID('seq3'), [],
graph.SequenceMetadata({'group': '2'}))
}
expected_poagraph = graph.Poagraph(expected_nodes, expected_sequences)
actual_poagraph, _ = builder.build_from_dagmaf(
msa.Maf(pathtools.get_file_content_stringio(maf_path), maf_path),
missings.ConstBaseProvider(self.missing_n), self.metadatacsv)
self.assertEqual(expected_poagraph, actual_poagraph)
def test_9_get_seqids(self):
metadata_path = self.csv_files_dir.joinpath("test_1_correct.csv")
csv_content = pathtools.get_file_content_stringio(metadata_path)
expected_seqids = [
msa.SequenceID('s1'),
msa.SequenceID('s2'),
msa.SequenceID('s3')
]
m = msa.MetadataCSV(csv_content, metadata_path)
actual_seqids = m.get_all_sequences_ids()
self.assertEqual(expected_seqids, actual_seqids)
def test_05_single_block_single_nucletodide(self):
maf_path = self.maf_files_dir.joinpath(
"test_5_single_block_single_nucletodide.maf")
expected_nodes = [
graph.Node(node_id=nid(0), base=graph.Base('A'), aligned_to=None, block_id=bid(0))
]
expected_sequences = {
msa.SequenceID('seq0'):
graph.Sequence(msa.SequenceID('seq0'),
[graph.SeqPath([*map(nid, [0])])],
graph.SequenceMetadata({'group': '1'})),
msa.SequenceID('seq1'):
graph.Sequence(msa.SequenceID('seq1'),
[graph.SeqPath([*map(nid, [0])])],
graph.SequenceMetadata({'group': '1'})),
msa.SequenceID('seq2'):
graph.Sequence(msa.SequenceID('seq2'),
[graph.SeqPath([*map(nid, [0])])],
graph.SequenceMetadata({'group': '2'})),
msa.SequenceID('seq3'):
graph.Sequence(msa.SequenceID('seq3'),
[graph.SeqPath([*map(nid, [0])])],
graph.SequenceMetadata({'group': '2'}))
}
expected_poagraph = graph.Poagraph(expected_nodes, expected_sequences)
actual_poagraph, _ = builder.build_from_dagmaf(
msa.Maf(pathtools.get_file_content_stringio(maf_path), maf_path),
self.fasta_provider,
self.metadatacsv)
self.assertEqual(expected_poagraph, actual_poagraph)
def test_2_three_sequences_in_two_files_in_zip(self):
fasta_path = self.fasta_dir.joinpath(
"test_2_three_sequences_in_two_files_in_zip.zip")
fasta_provider = missings.FromFile(Path(fasta_path))
sequence_id_1 = msa.SequenceID("seq1")
self.raise_error_if_unequal(sequence_id_1, "ACTGGGTGGGA",
fasta_provider)
sequence_id_2 = msa.SequenceID("seq2")
self.raise_error_if_unequal(sequence_id_2, "AA", fasta_provider)
sequence_id_3 = msa.SequenceID("seq3")
self.raise_error_if_unequal(sequence_id_3, "GT", fasta_provider)
def _get_children_nodes_looping(node: tree.AffinityNode,
poagraph: graph.Poagraph,
output_dir: Path,
blosum_path: Path,
p: parameters.P,
current_max_affinity_node_id: int) -> List[tree.AffinityNode]:
"""Generates children of given Affinity Tree node."""
children_nodes: List[tree.AffinityNode] = []
not_assigned_sequences_ids: List[msa.SequenceID] = node.sequences
detailed_logger.info(f"""Getting children nodes for
affinity node {node.id_}...""")
affinity_node_id = 0
so_far_cutoffs: List[poagraph.Compatibility] = []
while not_assigned_sequences_ids:
detailed_logger.info(f"### Getting child {len(so_far_cutoffs)}...")
child_ready = False
attempt = 0
current_candidates = not_assigned_sequences_ids
while not child_ready:
consensus_candidate = poa.get_consensuses(poagraph,
current_candidates,
output_dir,
f"parent_{node.id_}_child_{len(so_far_cutoffs)}_attempt_{attempt}",
blosum_path,
parameters.Hbmin(0),
specific_consensuses_id=[0])[0].path
compatibilities_to_consensus_candidate = poagraph.get_compatibilities(sequences_ids=not_assigned_sequences_ids,
consensus_path=consensus_candidate,
p=p)
compatibilities_to_consensus_candidate[msa.SequenceID("parent")] = node.mincomp
qualified_sequences_ids_candidates, cutoff = _get_qualified_sequences_ids_and_cutoff(
compatibilities_to_max_c=compatibilities_to_consensus_candidate,
so_far_cutoffs=so_far_cutoffs,
splitted_node_id=node.id_)
if qualified_sequences_ids_candidates == current_candidates or attempt == 10:
if attempt == 10:
detailed_logger.info("Attempt treshold 10 exceeded!")
affinity_node_id += 1
affinity_node = tree.AffinityNode(
id_=tree.AffinityNodeID(current_max_affinity_node_id + affinity_node_id),
parent=node.id_,
sequences=qualified_sequences_ids_candidates,
mincomp=_get_min_comp(node_sequences_ids=qualified_sequences_ids_candidates,
comps_to_consensus=compatibilities_to_consensus_candidate),
consensus=graph.SeqPath(consensus_candidate))
children_nodes.append(affinity_node)
not_assigned_sequences_ids = list(set(not_assigned_sequences_ids) - set(qualified_sequences_ids_candidates))
child_ready = True
so_far_cutoffs.append(affinity_node.mincomp)
else:
current_candidates = qualified_sequences_ids_candidates
attempt += 1
detailed_logger.info("Children nodes generated.")
return children_nodes
def test_1_download_sequence_and_save_to_cache(self):
cache_dir_path = pathtools.get_child_path(Path.cwd(), ".fastacache")
if cache_dir_path.exists():
shutil.rmtree(cache_dir_path)
ncbi_fasta_provider = missings.FromNCBI(use_cache=True)
sequence_id = msa.SequenceID("AB050936v1")
_ = ncbi_fasta_provider.get_base(sequence_id, 0)
# cache directory creation
cache_directory_created = cache_dir_path.exists()
self.assertTrue(cache_directory_created)
# file creation
files_in_cache_dircetory = [*cache_dir_path.glob("*")]
expected_filepath = pathtools.get_child_path(cache_dir_path,
f"{sequence_id}.fasta")
file_created_in_cache = expected_filepath in files_in_cache_dircetory
self.assertTrue(file_created_in_cache)
# file content
control_fasta_path = Path(__file__).parent.joinpath(
'fasta_ncbi/AB050936.1.fasta').resolve()
with open(control_fasta_path) as fasta_file_hanlder:
expected_content = fasta_file_hanlder.read()
with open(expected_filepath) as fasta_file_handler:
actual_content = fasta_file_handler.read()
self.assertEqual(expected_content, actual_content)
def test_1_no_symbol_provided(self):
missing_symbol = missings.MissingBase()
const_symbol_provider = missings.ConstBaseProvider(missing_symbol)
expected_symbol = graph.Base('?')
actual_symbol = const_symbol_provider.get_base(msa.SequenceID('s'), 0)
self.assertEqual(expected_symbol, actual_symbol)
def test_10_parallel_blocks_1st_and_2nd_merge_into_3rd(self):
maf_path = self.maf_files_dir.joinpath("test_10_parallel_blocks_1st_and_2nd_merge_into_3rd.maf")
expected_nodes = [
graph.Node(node_id=nid(0), base=graph.Base('G'), aligned_to=nid(1)),
graph.Node(node_id=nid(1), base=graph.Base('T'), aligned_to=nid(0)),
graph.Node(node_id=nid(2), base=graph.Base('T'), aligned_to=None),
graph.Node(node_id=nid(3), base=graph.Base('A'), aligned_to=None),
graph.Node(node_id=nid(4), base=graph.Base('C'), aligned_to=nid(5)),
graph.Node(node_id=nid(5), base=graph.Base('G'), aligned_to=nid(4)),
graph.Node(node_id=nid(6), base=graph.Base('C'), aligned_to=None),
graph.Node(node_id=nid(7), base=graph.Base('A'), aligned_to=None),
graph.Node(node_id=nid(8), base=graph.Base('C'), aligned_to=None),
graph.Node(node_id=nid(9), base=graph.Base('T'), aligned_to=None),
graph.Node(node_id=nid(10), base=graph.Base('G'), aligned_to=None),
graph.Node(node_id=nid(11), base=graph.Base('G'), aligned_to=None),
graph.Node(node_id=nid(12), base=graph.Base('C'), aligned_to=nid(13)),
graph.Node(node_id=nid(13), base=graph.Base('G'), aligned_to=nid(12)),
graph.Node(node_id=nid(14), base=graph.Base('C'), aligned_to=nid(15)),
graph.Node(node_id=nid(15), base=graph.Base('G'), aligned_to=nid(16)),
graph.Node(node_id=nid(16), base=graph.Base('T'), aligned_to=nid(14)),
graph.Node(node_id=nid(17), base=graph.Base('A'), aligned_to=nid(18)),
graph.Node(node_id=nid(18), base=graph.Base('T'), aligned_to=nid(17)),
graph.Node(node_id=nid(19), base=graph.Base('A'), aligned_to=nid(20)),
graph.Node(node_id=nid(20), base=graph.Base('C'), aligned_to=nid(19)),
graph.Node(node_id=nid(21), base=graph.Base('C'), aligned_to=nid(22)),
graph.Node(node_id=nid(22), base=graph.Base('G'), aligned_to=nid(21)),
]
expected_sequences = {
msa.SequenceID('seq0'):
graph.Sequence(msa.SequenceID('seq0'),
[graph.SeqPath([*map(nid, [7, 8, 9, 10, 11, 12, 15, 18, 19, 21])])],
graph.SequenceMetadata({'group': '1'})),
msa.SequenceID('seq1'):
graph.Sequence(msa.SequenceID('seq1'),
[graph.SeqPath([*map(nid, [7, 8, 9, 10, 11, 12, 15, 18, 19, 21])])],
graph.SequenceMetadata({'group': '1'})),
msa.SequenceID('seq2'):
graph.Sequence(msa.SequenceID('seq2'),
[graph.SeqPath([*map(nid, [0, 2, 3, 4, 6, 13, 16, 17, 20, 21])])],
graph.SequenceMetadata({'group': '2'})),
msa.SequenceID('seq3'):
graph.Sequence(msa.SequenceID('seq3'),
[graph.SeqPath([*map(nid, [1, 2, 3, 5, 6, 13, 14, 17, 20, 22])])],
graph.SequenceMetadata({'group': '2'})),
}
expected_poagraph = graph.Poagraph(expected_nodes, expected_sequences)
actual_poagraph, _ = builder.build_from_dagmaf(
msa.Maf(pathtools.get_file_content_stringio(maf_path), maf_path),
self.fasta_provider,
self.metadatacsv)
self.assertEqual(expected_poagraph, actual_poagraph)
def test_09_inactive_edges_but_all_strands_plus(self):
maf_path = self.maf_files_dir.joinpath("test_9_inactive_edges_but_all_strands_plus.maf")
expected_nodes = [
graph.Node(node_id=nid(0), base=graph.Base('A'), aligned_to=None),
graph.Node(node_id=nid(1), base=graph.Base('C'), aligned_to=None),
graph.Node(node_id=nid(2), base=graph.Base('T'), aligned_to=None),
graph.Node(node_id=nid(3), base=graph.Base('G'), aligned_to=None),
graph.Node(node_id=nid(4), base=graph.Base('G'), aligned_to=None),
graph.Node(node_id=nid(5), base=graph.Base('A'), aligned_to=None),
graph.Node(node_id=nid(6), base=graph.Base('C'), aligned_to=None),
graph.Node(node_id=nid(7), base=graph.Base('T'), aligned_to=None),
graph.Node(node_id=nid(8), base=graph.Base('G'), aligned_to=None),
graph.Node(node_id=nid(9), base=graph.Base('G'), aligned_to=None),
graph.Node(node_id=nid(10), base=graph.Base('A'), aligned_to=None),
graph.Node(node_id=nid(11), base=graph.Base('C'), aligned_to=None),
graph.Node(node_id=nid(12), base=graph.Base('T'), aligned_to=None),
graph.Node(node_id=nid(13), base=graph.Base('G'), aligned_to=None),
graph.Node(node_id=nid(14), base=graph.Base('G'), aligned_to=None),
graph.Node(node_id=nid(15), base=graph.Base('A'), aligned_to=None),
graph.Node(node_id=nid(16), base=graph.Base('C'), aligned_to=None),
graph.Node(node_id=nid(17), base=graph.Base('T'), aligned_to=None),
graph.Node(node_id=nid(18), base=graph.Base('G'), aligned_to=None),
graph.Node(node_id=nid(19), base=graph.Base('G'), aligned_to=None),
]
expected_sequences = {
msa.SequenceID('seq0'):
graph.Sequence(msa.SequenceID('seq0'),
[],
graph.SequenceMetadata({'group': '1'})),
msa.SequenceID('seq1'):
graph.Sequence(msa.SequenceID('seq1'),
[graph.SeqPath([*map(nid, [0, 1, 2, 3, 4, 10, 11, 12, 13, 14])]),
graph.SeqPath([*map(nid, [5, 6, 7, 8, 9, 15, 16, 17, 18, 19])])],
graph.SequenceMetadata({'group': '1'})),
msa.SequenceID('seq2'):
graph.Sequence(msa.SequenceID('seq2'),
[graph.SeqPath([*map(nid, [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12,
13, 14, 15, 16, 17, 18, 19])])],
graph.SequenceMetadata({'group': '2'})),
msa.SequenceID('seq3'):
graph.Sequence(msa.SequenceID('seq3'),
[],
graph.SequenceMetadata({'group': '2'})),
}
expected_poagraph = graph.Poagraph(expected_nodes, expected_sequences)
actual_poagraph, _ = builder.build_from_dagmaf(
msa.Maf(pathtools.get_file_content_stringio(maf_path), maf_path),
self.fasta_provider,
self.metadatacsv)
self.assertEqual(expected_poagraph, actual_poagraph)
def test_6_missing_one_reverted_sequence_middle_minus1_1(self):
maf_path = self.maf_files_dir.joinpath(
"test_6_missing_one_reverted_sequence_middle_minus1_1.maf")
expected_nodes = [
# block 1 because it is first in DAG and reverted
graph.Node(node_id=nid(0), base=graph.Base('A'), aligned_to=None),
graph.Node(node_id=nid(1), base=graph.Base('C'), aligned_to=None),
graph.Node(node_id=nid(2), base=graph.Base('C'),
aligned_to=nid(3)),
graph.Node(node_id=nid(3), base=graph.Base('T'),
aligned_to=nid(2)),
# missing seq2, on edge (-1,1)
graph.Node(node_id=nid(4), base=graph.Base('T'), aligned_to=None),
graph.Node(node_id=nid(5), base=graph.Base('C'), aligned_to=None),
graph.Node(node_id=nid(6), base=graph.Base('A'),
aligned_to=nid(7)),
graph.Node(node_id=nid(7), base=graph.Base('C'),
aligned_to=nid(6)),
graph.Node(node_id=nid(8), base=graph.Base('C'), aligned_to=None),
graph.Node(node_id=nid(9), base=graph.Base('T'), aligned_to=None),
graph.Node(node_id=nid(10),
base=graph.Base('A'),
aligned_to=nid(11)),
graph.Node(node_id=nid(11),
base=graph.Base('C'),
aligned_to=nid(10)),
]
expected_sequences = {
msa.SequenceID('seq0'):
graph.Sequence(msa.SequenceID('seq0'), [],
graph.SequenceMetadata({'group': '1'})),
msa.SequenceID('seq1'):
graph.Sequence(msa.SequenceID('seq1'), [
graph.SeqPath([*map(nid, [0, 1, 2])]),
graph.SeqPath([*map(nid, [6, 8, 9, 10])])
], graph.SequenceMetadata({'group': '1'})),
msa.SequenceID('seq2'):
graph.Sequence(
msa.SequenceID('seq2'),
[graph.SeqPath([*map(nid, [0, 1, 3, 4, 5, 7, 11])])],
graph.SequenceMetadata({'group': '2'})),
msa.SequenceID('seq3'):
graph.Sequence(msa.SequenceID('seq3'), [],
graph.SequenceMetadata({'group': '2'}))
}
expected_poagraph = graph.Poagraph(expected_nodes, expected_sequences)
actual_poagraph, _ = builder.build_from_dagmaf(
msa.Maf(pathtools.get_file_content_stringio(maf_path), maf_path),
self.fasta_provider, self.metadatacsv)
self.assertEqual(expected_poagraph, actual_poagraph)
def _add_record_to_dict(self, record: SeqRecord,
fastas_dict: Dict[msa.SequenceID, str]) -> None:
"""Adds record the sequence to dict if the sequence is not empty.
Args:
record: The sequence to add.
Raises:
FastaProviderExecption: If record contains empty sequence or
its ID is already present in the dict.
"""
if len(record.seq) == 0:
raise FastaProviderException(
"Empty sequence in FASTA. Provide the sequence or remove its header."
)
if msa.SequenceID(str(record.id)) in fastas_dict.keys():
raise FastaProviderException(
"Incorrect fasta provided: sequences IDs are not unique.")
fastas_dict[msa.SequenceID(str(record.id))] = record.seq
def test_1_one_sequence(self):
fasta_path = self.fasta_dir.joinpath("test_1_one_sequence.fasta")
fasta_provider = missings.FromFile(Path(fasta_path))
sequence_id = msa.SequenceID("seq1")
expected_sequence = self.read_sequence(fasta_path)
self.raise_error_if_unequal(sequence_id,
expected_sequence,
fasta_provider)
def get_poagraph(maf: msa.Maf, metadata: Optional[msa.MetadataCSV]) -> \
Tuple[List[graph.Node], Dict[msa.SequenceID, graph.Sequence]]:
"""Get poagraph elements from MAF.
Args:
maf: Multialignment file in MAF format.
metadata: MetadataCSV.
Returns:
Tuple of poagraph elements.
"""
alignment = [*AlignIO.parse(maf.filecontent, "maf")]
nodes, sequences = _init_poagraph(alignment, metadata)
current_node_id = graph.NodeID(-1)
column_id = graph.ColumnID(-1)
for block_id, block in enumerate(alignment):
global_logger.info(f"Processing block {block_id}...")
block_width = len(block[0].seq)
for col in range(block_width):
column_id += 1
sequence_id_to_nucleotide = {
msa.SequenceID(seq.id): seq[col]
for seq in block
}
nodes_codes = sorted([
*(set([
nucleotide
for nucleotide in sequence_id_to_nucleotide.values()
])).difference({'-'})
])
column_nodes_ids = [
graph.NodeID(current_node_id + i + 1)
for i, _ in enumerate(nodes_codes)
]
for i, nucl in enumerate(nodes_codes):
current_node_id += 1
nodes.append(
graph.Node(node_id=current_node_id,
base=graph.Base(nucl),
aligned_to=_get_next_aligned_node_id(
graph.NodeID(i), column_nodes_ids),
column_id=graph.ColumnID(column_id),
block_id=graph.BlockID(block_id)))
for seq_id, nucleotide in sequence_id_to_nucleotide.items():
if nucleotide == nucl:
sequences[seq_id] = _add_node_do_sequence(
sequence=sequences[seq_id],
node_id=current_node_id)
return nodes, sequences
def _get_paths_join_info(block: Block,
free_edges: Dict[msa.SequenceID, List[Edge]]) -> \
Dict[msa.SequenceID, Optional[graph.NodeID]]:
paths_join_info: Dict[msa.SequenceID, Optional[graph.NodeID]] = dict()
for seq in block.alignment:
seq_id = msa.SequenceID(seq.id)
paths_join_info[seq_id] = None
for i, edge in enumerate(free_edges[seq_id]):
if edge.to_block_id == block.id:
paths_join_info[seq_id] = edge.last_node_id
return paths_join_info
def test_1_one_sequence_one_file_in_zip(self):
fasta_path = self.fasta_dir.joinpath(
"test_1_one_sequence_one_file_in_zip.zip")
fasta_provider = missings.FromFile(Path(fasta_path))
sequence_id = msa.SequenceID("seq1")
expected_sequence = "ACTGGGTGGGA"
self.raise_error_if_unequal(sequence_id, expected_sequence,
fasta_provider)
def test_1_typical_poagraph(self):
po_path = self.po_files_dir.joinpath("test_1.po")
expected_nodes = [
graph.Node(node_id=nid(0), base=bid('A'), aligned_to=nid(1)),
graph.Node(node_id=nid(1), base=bid('G'), aligned_to=nid(0)),
graph.Node(node_id=nid(2), base=bid('C'), aligned_to=nid(3)),
graph.Node(node_id=nid(3), base=bid('G'), aligned_to=nid(2)),
graph.Node(node_id=nid(4), base=bid('A'), aligned_to=nid(5)),
graph.Node(node_id=nid(5), base=bid('T'), aligned_to=nid(4)),
graph.Node(node_id=nid(6), base=bid('G'), aligned_to=None),
graph.Node(node_id=nid(7), base=bid('G'), aligned_to=None),
graph.Node(node_id=nid(8), base=bid('A'), aligned_to=nid(9)),
graph.Node(node_id=nid(9), base=bid('C'), aligned_to=nid(10)),
graph.Node(node_id=nid(10), base=bid('G'), aligned_to=nid(11)),
graph.Node(node_id=nid(11), base=bid('T'), aligned_to=nid(8)),
graph.Node(node_id=nid(12), base=bid('A'), aligned_to=nid(13)),
graph.Node(node_id=nid(13), base=bid('C'), aligned_to=nid(12)),
graph.Node(node_id=nid(14), base=bid('T'), aligned_to=None),
graph.Node(node_id=nid(15), base=bid('A'), aligned_to=nid(16)),
graph.Node(node_id=nid(16), base=bid('C'), aligned_to=nid(17)),
graph.Node(node_id=nid(17), base=bid('G'), aligned_to=nid(15))
]
expected_sequences = {
msa.SequenceID('seq0'):
graph.Sequence(
msa.SequenceID('seq0'),
[graph.SeqPath([*map(nid, [0, 2, 4, 6, 7, 8, 12, 14, 16])])],
graph.SequenceMetadata({'group': '1'})),
msa.SequenceID('seq1'):
graph.Sequence(msa.SequenceID('seq1'),
[graph.SeqPath([*map(nid, [1, 2, 5, 6, 7, 9])])],
graph.SequenceMetadata({'group': '1'})),
msa.SequenceID('seq2'):
graph.Sequence(
msa.SequenceID('seq2'),
[graph.SeqPath([*map(nid, [3, 4, 6, 7, 10, 12, 14, 17])])],
graph.SequenceMetadata({'group': '2'})),
msa.SequenceID('seq3'):
graph.Sequence(msa.SequenceID('seq3'),
[graph.SeqPath([*map(nid, [11, 13, 14, 15])])],
graph.SequenceMetadata({'group': '2'}))
}
expected_poagraph = graph.Poagraph(expected_nodes, expected_sequences)
nodes, sequences = po2poagraph.get_poagraph(
msa.Po(pathtools.get_file_content_stringio(po_path), po_path),
self.metadatacsv)
actual_poagraph = graph.Poagraph(nodes, sequences)
self.assertEqual(expected_poagraph, actual_poagraph)
def test_2_missing_sequence_end(self):
maf_path = self.maf_files_dir.joinpath(
"test_2_missing_sequence_end.maf")
expected_nodes = [
graph.Node(node_id=nid(0), base=graph.Base('A'),
aligned_to=nid(1)),
graph.Node(node_id=nid(1), base=graph.Base('G'),
aligned_to=nid(0)),
graph.Node(node_id=nid(2), base=graph.Base('C'),
aligned_to=nid(3)),
graph.Node(node_id=nid(3), base=graph.Base('G'),
aligned_to=nid(2)),
graph.Node(node_id=nid(4), base=graph.Base('T'), aligned_to=None),
graph.Node(node_id=nid(5), base=graph.Base('A'),
aligned_to=nid(6)),
graph.Node(node_id=nid(6), base=graph.Base('C'),
aligned_to=nid(5)),
graph.Node(node_id=nid(7), base=graph.Base('A'), aligned_to=None),
graph.Node(node_id=nid(8), base=graph.Base('G'), aligned_to=None),
graph.Node(node_id=nid(9), base=graph.Base('G'), aligned_to=None),
graph.Node(node_id=nid(10), base=graph.Base('T'), aligned_to=None),
graph.Node(node_id=nid(11), base=graph.Base('G'), aligned_to=None),
graph.Node(node_id=nid(12), base=graph.Base('T'), aligned_to=None),
]
expected_sequences = {
msa.SequenceID('seq0'):
graph.Sequence(msa.SequenceID('seq0'), [],
graph.SequenceMetadata({'group': '1'})),
msa.SequenceID('seq1'):
graph.Sequence(
msa.SequenceID('seq1'),
[graph.SeqPath([*map(nid, [0, 2, 4, 5, 8, 9, 10])])],
graph.SequenceMetadata({'group': '1'})),
msa.SequenceID('seq2'):
graph.Sequence(
msa.SequenceID('seq2'),
[graph.SeqPath([*map(nid, [1, 3, 4, 6, 7, 11, 12])])],
graph.SequenceMetadata({'group': '2'})),
msa.SequenceID('seq3'):
graph.Sequence(msa.SequenceID('seq3'), [],
graph.SequenceMetadata({'group': '2'}))
}
expected_poagraph = graph.Poagraph(expected_nodes, expected_sequences)
actual_poagraph, _ = builder.build_from_dagmaf(
msa.Maf(pathtools.get_file_content_stringio(maf_path), maf_path),
self.fasta_provider, self.metadatacsv)
self.assertEqual(expected_poagraph, actual_poagraph)
def test_4_seqid_is_last(self):
metadata_path = self.csv_files_dir.joinpath("test_4_seqid_is_last.csv")
csv_content = pathtools.get_file_content_stringio(metadata_path)
expected_metadata = {
msa.SequenceID('s1'): {
'name': 'sequence1',
'group': 'A'
},
msa.SequenceID('s2'): {
'name': 'sequence2',
'group': 'B'
},
msa.SequenceID('s3'): {
'name': 'sequence3',
'group': 'B'
}
}
m = msa.MetadataCSV(csv_content, metadata_path)
actual_metadata = m.metadata
self.assertEqual(expected_metadata, actual_metadata)
def test_1_typical_poagraph(self):
expected_po_content_path = self.po_files_dir.joinpath("test_1.po")
poagraph_nodes = [graph.Node(node_id=nid(0), base=bid('A'), aligned_to=nid(1)),
graph.Node(node_id=nid(1), base=bid('G'), aligned_to=nid(0)),
graph.Node(node_id=nid(2), base=bid('C'), aligned_to=nid(3)),
graph.Node(node_id=nid(3), base=bid('G'), aligned_to=nid(2)),
graph.Node(node_id=nid(4), base=bid('A'), aligned_to=nid(5)),
graph.Node(node_id=nid(5), base=bid('T'), aligned_to=nid(4)),
graph.Node(node_id=nid(6), base=bid('G'), aligned_to=None),
graph.Node(node_id=nid(7), base=bid('G'), aligned_to=None),
graph.Node(node_id=nid(8), base=bid('A'), aligned_to=nid(9)),
graph.Node(node_id=nid(9), base=bid('C'), aligned_to=nid(10)),
graph.Node(node_id=nid(10), base=bid('G'), aligned_to=nid(11)),
graph.Node(node_id=nid(11), base=bid('T'), aligned_to=nid(8)),
graph.Node(node_id=nid(12), base=bid('A'), aligned_to=nid(13)),
graph.Node(node_id=nid(13), base=bid('C'), aligned_to=nid(12)),
graph.Node(node_id=nid(14), base=bid('T'), aligned_to=None),
graph.Node(node_id=nid(15), base=bid('A'), aligned_to=nid(16)),
graph.Node(node_id=nid(16), base=bid('C'), aligned_to=nid(17)),
graph.Node(node_id=nid(17), base=bid('G'), aligned_to=nid(15))
]
poagraph_sequences = {
msa.SequenceID('seq0'):
graph.Sequence(msa.SequenceID('seq0'),
[graph.SeqPath([*map(nid, [0, 2, 4, 6, 7, 8, 12, 14, 16])])],
graph.SequenceMetadata({'group': '1'})),
msa.SequenceID('seq1'):
graph.Sequence(msa.SequenceID('seq1'),
[graph.SeqPath([*map(nid, [1, 2, 5, 6, 7, 9])])],
graph.SequenceMetadata({'group': '1'})),
msa.SequenceID('seq2'):
graph.Sequence(msa.SequenceID('seq2'),
[graph.SeqPath([*map(nid, [3, 4, 6, 7, 10, 12, 14, 17])])],
graph.SequenceMetadata({'group': '1'})),
msa.SequenceID('seq3'):
graph.Sequence(msa.SequenceID('seq3'),
[graph.SeqPath([*map(nid, [11, 13, 14, 15])])],
graph.SequenceMetadata({'group': '1'})),
}
poagraph = graph.Poagraph(poagraph_nodes, poagraph_sequences)
actual_po_content = po.poagraph_to_PangenomePO(poagraph)
expected_po_content = pathtools.get_file_content(expected_po_content_path)
self.assertEqual(expected_po_content, actual_po_content)
def test_subpoagraph_construction_from_poagraph_keep_seq_0_1(self):
translator = poa._PoagraphPOTranslator(
self.poagraph, [msa.SequenceID('seq0'),
msa.SequenceID('seq1')])
actual_po_content = translator.get_input_po_content()
expected_po_content = "VERSION=pangenome\n"\
"NAME=pangenome\n"\
"TITLE=pangenome\n"\
"LENGTH=9\n"\
"SOURCECOUNT=2\n"\
"SOURCENAME=seq0\n"\
"SOURCEINFO=7 0 0 -1 seq0\n"\
"SOURCENAME=seq1\n"\
"SOURCEINFO=5 1 100 -1 seq1\n"\
"t:S0\n"\
"a:L0S0S1\n"\
"a:L1S0S1\n"\
"a:L2S0A4\n"\
"c:L2S1A3\n"\
"a:L3L4S0S1\n"\
"c:L5S0A7\n"\
"t:L5S1A6\n"\
"a:L6S0"
self.assertEqual(expected_po_content, actual_po_content)
def _complement_sequence_middles_if_needed(build_state: _BuildState,
block: Block, edge: Arc, seq,
last_node_id: graph.NodeID):
seq_id = msa.SequenceID(seq[0].seq_id)
left_block_sinfo, right_block_sinfo = _get_edge_sinfos(
seqs_info=build_state.seqs_info,
from_block_id=block.id,
edge=edge,
seq_id=seq_id)
if _complementation_not_needed(left_block_sinfo, right_block_sinfo):
if edge.edge_type == (1, -1):
return last_node_id
else:
return None
else:
current_node_id = _get_max_node_id(build_state.nodes)
column_id = build_state.column_id
if left_block_sinfo.start < right_block_sinfo.start:
last_pos = left_block_sinfo.start + left_block_sinfo.size - 1
next_pos = right_block_sinfo.start
else:
last_pos = right_block_sinfo.start + right_block_sinfo.size - 1
next_pos = left_block_sinfo.start
join_with = last_node_id if _should_join_with_last_node(
edge.edge_type) else None
for i in range(last_pos + 1, next_pos):
column_id += 1
current_node_id += 1
missing_nucleotide = _get_missing_nucleotide(
build_state.fasta_provider, seq_id, i)
build_state.nodes += [
graph.Node(node_id=current_node_id,
base=missing_nucleotide,
aligned_to=None,
column_id=column_id,
block_id=None)
]
_add_node_to_sequence(build_state,
seq_id=seq_id,
join_with=join_with,
node_id=current_node_id)
join_with = current_node_id
if _should_join_with_next_node(edge.edge_type):
return current_node_id
else:
return None
def setUp(self):
nodes = [
graph.Node(
node_id=nid(0),
base=b('T'),
aligned_to=None,
),
graph.Node(node_id=nid(1), base=b('A'), aligned_to=nid(2)),
graph.Node(node_id=nid(2), base=b('G'), aligned_to=nid(1)),
graph.Node(node_id=nid(3), base=b('A'), aligned_to=nid(4)),
graph.Node(node_id=nid(4), base=b('C'), aligned_to=nid(3)),
graph.Node(node_id=nid(5), base=b('A'), aligned_to=nid(6)),
graph.Node(node_id=nid(6), base=b('C'), aligned_to=nid(7)),
graph.Node(node_id=nid(7), base=b('G'), aligned_to=nid(8)),
graph.Node(node_id=nid(8), base=b('T'), aligned_to=nid(5)),
graph.Node(node_id=nid(9), base=b('A'), aligned_to=None),
graph.Node(node_id=nid(10), base=b('C'), aligned_to=nid(11)),
graph.Node(node_id=nid(11), base=b('T'), aligned_to=nid(10)),
graph.Node(node_id=nid(12), base=b('G'), aligned_to=None),
graph.Node(node_id=nid(13), base=b('A'), aligned_to=nid(14)),
graph.Node(node_id=nid(14), base=b('C'), aligned_to=nid(13))
]
sequences = {
msa.SequenceID('seq0'):
graph.Sequence(
msa.SequenceID('seq0'),
[graph.SeqPath([*map(nid, [0, 1, 3, 5, 9, 10, 13])])],
graph.SequenceMetadata({'group': '1'})),
msa.SequenceID('seq1'):
graph.Sequence(msa.SequenceID('seq1'),
[graph.SeqPath([*map(nid, [1, 3, 6, 9, 11])])],
graph.SequenceMetadata({'group': '1'})),
msa.SequenceID('seq2'):
graph.Sequence(msa.SequenceID('seq2'),
[graph.SeqPath([*map(nid, [2, 4, 7, 9, 11, 12])])],
graph.SequenceMetadata({'group': '1'})),
msa.SequenceID('seq3'):
graph.Sequence(
msa.SequenceID('seq3'),
[graph.SeqPath([*map(nid, [2, 4, 8, 9, 11, 12, 14])])],
graph.SequenceMetadata({'group': '1'})),
}
self.poagraph = graph.Poagraph(nodes, sequences)
def test_read_consensus_path_seq1_only_in_input(self):
translator = poa._PoagraphPOTranslator(self.poagraph,
[msa.SequenceID('seq1')])
_ = translator.get_input_po_content()
poa_lines = [
"VERSION=pangenome\n", "NAME=pangenome\n", "TITLE=pangenome\n",
"LENGTH=5\n", "SOURCECOUNT=2\n", "SOURCENAME=seq1\n",
"SOURCEINFO=5 0 100 0 seq1\n", "SOURCENAME=CONSENS0\n",
"SOURCEINFO=5 0 100 0 CONSENS0\n", "a:S0S1\n", "a:L0S0S1\n",
"c:L1S0S1\n", "a:L2S0S1\n", "t:L2S0S1"
]
actual_consensus_path = translator.read_consensus_paths(poa_lines, [0])
expected_consensus_path = [1, 3, 6, 9, 11]
self.assertEqual(expected_consensus_path,
actual_consensus_path[0].path)
def test_06_1st_block_separates_into_2_branches_which_connect_in_3rd_block(self):
maf_path = self.maf_files_dir.joinpath(
"test_6_1st_block_separates_into_2_branches_which_connect_in_3rd_block.maf")
expected_nodes = [
graph.Node(node_id=nid(0), base=graph.Base('A'), aligned_to=nid(1)),
graph.Node(node_id=nid(1), base=graph.Base('C'), aligned_to=nid(2)),
graph.Node(node_id=nid(2), base=graph.Base('G'), aligned_to=nid(0)),
graph.Node(node_id=nid(3), base=graph.Base('C'), aligned_to=None),
graph.Node(node_id=nid(4), base=graph.Base('A'), aligned_to=nid(5)),
graph.Node(node_id=nid(5), base=graph.Base('T'), aligned_to=nid(4)),
graph.Node(node_id=nid(6), base=graph.Base('G'), aligned_to=None),
graph.Node(node_id=nid(7), base=graph.Base('G'), aligned_to=None),
graph.Node(node_id=nid(8), base=graph.Base('C'), aligned_to=nid(9)),
graph.Node(node_id=nid(9), base=graph.Base('G'), aligned_to=nid(10)),
graph.Node(node_id=nid(10), base=graph.Base('T'), aligned_to=nid(8)),
]
expected_sequences = {
msa.SequenceID('seq0'):
graph.Sequence(msa.SequenceID('seq0'),
[graph.SeqPath([*map(nid, [0, 3, 4, 8])])],
graph.SequenceMetadata({'group': '1'})),
msa.SequenceID('seq1'):
graph.Sequence(msa.SequenceID('seq1'),
[graph.SeqPath([*map(nid, [1, 3, 5, 6, 7, 9])])],
graph.SequenceMetadata({'group': '1'})),
msa.SequenceID('seq2'):
graph.Sequence(msa.SequenceID('seq2'),
[graph.SeqPath([*map(nid, [2, 3, 5, 10])])],
graph.SequenceMetadata({'group': '2'})),
msa.SequenceID('seq3'):
graph.Sequence(msa.SequenceID('seq3'),
[],
graph.SequenceMetadata({'group': '2'}))
}
expected_poagraph = graph.Poagraph(expected_nodes, expected_sequences)
actual_poagraph, _ = builder.build_from_dagmaf(
msa.Maf(pathtools.get_file_content_stringio(maf_path), maf_path),
self.fasta_provider,
self.metadatacsv)
self.assertEqual(expected_poagraph, actual_poagraph)
def test_02_seq_starts_in_second_block(self):
maf_path = self.maf_files_dir.joinpath(
"test_2_seq_starts_in_second_block.maf")
expected_nodes = [
graph.Node(node_id=nid(0), base=graph.Base('C'), aligned_to=None, block_id=bid(0)),
graph.Node(node_id=nid(1), base=graph.Base('T'), aligned_to=None, block_id=bid(0)),
graph.Node(node_id=nid(2), base=graph.Base('G'), aligned_to=None, block_id=bid(0)),
graph.Node(node_id=nid(3), base=graph.Base('T'), aligned_to=None, block_id=bid(1)),
graph.Node(node_id=nid(4), base=graph.Base('G'), aligned_to=nid(5), block_id=bid(2)),
graph.Node(node_id=nid(5), base=graph.Base('T'), aligned_to=nid(4), block_id=bid(2)),
graph.Node(node_id=nid(6), base=graph.Base('A'), aligned_to=None, block_id=bid(2)),
graph.Node(node_id=nid(7), base=graph.Base('A'), aligned_to=None, block_id=bid(2)),
graph.Node(node_id=nid(8), base=graph.Base('C'), aligned_to=None, block_id=bid(2)),
]
expected_sequences = {
msa.SequenceID('seq0'):
graph.Sequence(msa.SequenceID('seq0'),
[graph.SeqPath([*map(nid, [1, 2, 3])])],
graph.SequenceMetadata({'group': '1'})),
msa.SequenceID('seq1'):
graph.Sequence(msa.SequenceID('seq1'),
[graph.SeqPath([*map(nid, [0, 5, 7])])],
graph.SequenceMetadata({'group': '1'})),
msa.SequenceID('seq2'):
graph.Sequence(msa.SequenceID('seq2'),
[graph.SeqPath([*map(nid, [3, 4, 6, 8])])],
graph.SequenceMetadata({'group': '2'})),
msa.SequenceID('seq3'):
graph.Sequence(msa.SequenceID('seq3'),
[],
graph.SequenceMetadata({'group': '2'}))
}
expected_poagraph = graph.Poagraph(expected_nodes, expected_sequences)
actual_poagraph, _ = builder.build_from_dagmaf(
msa.Maf(pathtools.get_file_content_stringio(maf_path), maf_path),
self.fasta_provider,
self.metadatacsv)
self.assertEqual(expected_poagraph, actual_poagraph)