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 _add_node_do_sequence(sequence: graph.Sequence, node_id: graph.NodeID) -> \
graph.Sequence:
if sequence.paths:
a = graph.SeqPath([node_id])
updated_path = graph.SeqPath(sequence.paths[-1] + a)
newpaths = [sequence.paths[:-1] + updated_path]
else:
newpaths = [graph.SeqPath([node_id])]
return graph.Sequence(sequence.seqid, newpaths, sequence.seqmetadata)
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)
def test_08_reversed_block(self):
maf_path = self.maf_files_dir.joinpath("test_8_reversed_block.maf")
expected_nodes = [
graph.Node(node_id=nid(0), base=graph.Base('C'), aligned_to=None),
graph.Node(node_id=nid(1), base=graph.Base('A'), aligned_to=None),
graph.Node(node_id=nid(2), base=graph.Base('T'), aligned_to=None),
# next block is reversed because it was converted to dag
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=nid(6)),
graph.Node(node_id=nid(6), base=graph.Base('G'), 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('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, 1, 3, 4, 5, 7, 8, 9])])],
graph.SequenceMetadata({'group': '1'})),
msa.SequenceID('seq2'):
graph.Sequence(msa.SequenceID('seq2'),
[graph.SeqPath([*map(nid, [0, 1, 2, 3, 4, 6, 7, 8, 9])])],
graph.SequenceMetadata({'group': '2'})),
msa.SequenceID('seq3'):
graph.Sequence(msa.SequenceID('seq3'),
[graph.SeqPath([*map(nid, [0, 1, 2, 3, 4, 6, 7, 9])])],
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 _init_sequences(sequences_info: Dict[int, POSequenceInfo],
metadata: Optional[msa.MetadataCSV]) -> \
Dict[msa.SequenceID, graph.Sequence]:
metadata_sequences_ids = metadata.get_all_sequences_ids() \
if metadata else []
po_sequences_ids = [seq_info.name for seq_info in sequences_info.values()]
initial_sequences = {seq_id: graph.Sequence(seqid=seq_id,
paths=[],
seqmetadata=metadata.get_sequence_metadata(seq_id)
if metadata else {})
for seq_id in set(po_sequences_ids + metadata_sequences_ids)}
return initial_sequences
def _init_sequences(sequences_in_dagmaf: List[msa.SequenceID],
metadata: Optional[msa.MetadataCSV]) -> \
Dict[msa.SequenceID, graph.Sequence]:
metadata_sequences_ids = metadata.get_all_sequences_ids(
) if metadata else []
initial_sequences = {
seq_id:
graph.Sequence(seqid=seq_id,
paths=[],
seqmetadata=metadata.get_sequence_metadata(seq_id)
if metadata else {})
for seq_id in set(sequences_in_dagmaf + metadata_sequences_ids)
}
return initial_sequences
def test_subpoagraph_should_omit_in_nodes_and_aligned_nodes(self):
# original poagraph
nodes = [
graph.Node(node_id=nid(0), base=b('A'), aligned_to=None),
graph.Node(node_id=nid(1), base=b('C'), aligned_to=nid(2)),
graph.Node(node_id=nid(2), base=b('T'), aligned_to=nid(1)),
graph.Node(node_id=nid(3), base=b('G'), aligned_to=None)
]
sequences = {
msa.SequenceID('seq1'):
graph.Sequence(msa.SequenceID('seq1'),
[graph.SeqPath([*map(nid, [0, 1, 3])])],
graph.SequenceMetadata({'group': '1'})),
msa.SequenceID('seq2'):
graph.Sequence(msa.SequenceID('seq2'),
[graph.SeqPath([*map(nid, [0, 2, 3])])],
graph.SequenceMetadata({'group': '1'}))
}
poagraph = graph.Poagraph(nodes, sequences)
translator = poa._PoagraphPOTranslator(poagraph,
[msa.SequenceID('seq2')])
actual_po_content = translator.get_input_po_content()
expected_po_content = "VERSION=pangenome\n" \
"NAME=pangenome\n" \
"TITLE=pangenome\n" \
"LENGTH=3\n" \
"SOURCECOUNT=1\n" \
"SOURCENAME=seq2\n" \
"SOURCEINFO=3 0 100 -1 seq2\n" \
"a:S0\n" \
"t:L0S0\n" \
"g:L1S0"
self.assertEqual(expected_po_content, actual_po_content)
def _init_poagraph(alignment: _ParsedMaf,
metadata: Optional[msa.MetadataCSV]) -> \
Tuple[List[graph.Node], Dict[msa.SequenceID, graph.Sequence]]:
maf_sequences_ids = _get_sequences_ids(alignment)
metadata_sequences_ids = metadata.get_all_sequences_ids(
) if metadata else []
initial_sequences = {
seq_id:
graph.Sequence(seqid=seq_id,
paths=[],
seqmetadata=metadata.get_sequence_metadata(seq_id)
if metadata else {})
for seq_id in set(maf_sequences_ids + metadata_sequences_ids)
}
return [], initial_sequences
def test_2_consensuses_and_empty_sequences(self):
po_path = self.po_files_dir.joinpath("test_2.po")
expected_nodes = [
graph.Node(node_id=nid(0), base=bid('C'), aligned_to=nid(1)),
graph.Node(node_id=nid(1), base=bid('T'), aligned_to=nid(0)),
graph.Node(node_id=nid(2), base=bid('A'), 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('C'), aligned_to=None),
graph.Node(node_id=nid(5), base=bid('T'), aligned_to=None),
graph.Node(node_id=nid(6), base=bid('A'), aligned_to=nid(7)),
graph.Node(node_id=nid(7), base=bid('T'), aligned_to=nid(6)),
graph.Node(node_id=nid(8), base=bid('G'), aligned_to=None)
]
expected_sequences = {
msa.SequenceID('seq0'):
graph.Sequence(msa.SequenceID('seq0'),
[graph.SeqPath([*map(nid, [0, 3, 4, 5, 6, 8])])],
graph.SequenceMetadata({'group': '1'})),
msa.SequenceID('seq1'):
graph.Sequence(msa.SequenceID('seq1'),
[graph.SeqPath([*map(nid, [1, 2, 4, 5, 7, 8])])],
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'})),
msa.SequenceID('CONSENS0'):
graph.Sequence(msa.SequenceID('CONSENS0'),
[graph.SeqPath([*map(nid, [0, 3, 4, 5, 7, 8])])],
graph.SequenceMetadata({})),
msa.SequenceID('CONSENS1'):
graph.Sequence(msa.SequenceID('CONSENS1'),
[graph.SeqPath([*map(nid, [1, 2, 4, 5, 6, 8])])],
graph.SequenceMetadata({}))
}
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_consensuses_and_empty_sequences(self):
expected_po_content_path = self.po_files_dir.joinpath("test_2.po")
poagraph_nodes = [graph.Node(node_id=nid(0), base=bid('C'), aligned_to=nid(1)),
graph.Node(node_id=nid(1), base=bid('T'), aligned_to=nid(0)),
graph.Node(node_id=nid(2), base=bid('A'), 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('C'), aligned_to=None),
graph.Node(node_id=nid(5), base=bid('T'), aligned_to=None),
graph.Node(node_id=nid(6), base=bid('A'), aligned_to=nid(7)),
graph.Node(node_id=nid(7), base=bid('T'), aligned_to=nid(6)),
graph.Node(node_id=nid(8), base=bid('G'), aligned_to=None)
]
poagraph_sequences = {
msa.SequenceID('seq0'):
graph.Sequence(msa.SequenceID('seq0'),
[graph.SeqPath([*map(nid, [0, 3, 4, 5, 6, 8])])],
graph.SequenceMetadata({'group': '1'})),
msa.SequenceID('seq1'):
graph.Sequence(msa.SequenceID('seq1'),
[graph.SeqPath([*map(nid, [1, 2, 4, 5, 7, 8])])],
graph.SequenceMetadata({'group': '1'})),
msa.SequenceID('seq2'):
graph.Sequence(msa.SequenceID('seq2'),
[],
graph.SequenceMetadata({'group': '1'})),
msa.SequenceID('seq3'):
graph.Sequence(msa.SequenceID('seq3'),
[],
graph.SequenceMetadata({'group': '1'})),
msa.SequenceID('CONSENS0'):
graph.Sequence(msa.SequenceID('CONSENS0'),
[graph.SeqPath([*map(nid, [0, 3, 4, 5, 7, 8])])],
None),
msa.SequenceID('CONSENS1'):
graph.Sequence(msa.SequenceID('CONSENS1'),
[graph.SeqPath([*map(nid, [1, 2, 4, 5, 6, 8])])],
None),
}
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_1_missing_sequence_start(self):
maf_path = self.maf_files_dir.joinpath(
"test_1_missing_sequence_start.maf")
expected_nodes = [
graph.Node(node_id=nid(0),
base=graph.Base(self.missing_n.value),
aligned_to=None,
block_id=bid(0)),
graph.Node(node_id=nid(1),
base=graph.Base(self.missing_n.value),
aligned_to=None,
block_id=bid(0)),
graph.Node(node_id=nid(2),
base=graph.Base(self.missing_n.value),
aligned_to=None,
block_id=bid(0)),
graph.Node(node_id=nid(3), base=graph.Base('A'),
aligned_to=nid(4)),
graph.Node(node_id=nid(4), base=graph.Base('G'),
aligned_to=nid(3)),
graph.Node(node_id=nid(5),
base=graph.Base('G'),
aligned_to=None,
block_id=bid(0)),
graph.Node(node_id=nid(6), base=graph.Base('G'),
aligned_to=nid(7)),
graph.Node(node_id=nid(7), base=graph.Base('T'),
aligned_to=nid(6)),
graph.Node(node_id=nid(8),
base=graph.Base('C'),
aligned_to=None,
block_id=bid(0)),
graph.Node(node_id=nid(9),
base=graph.Base('A'),
aligned_to=None,
block_id=bid(0)),
graph.Node(node_id=nid(10),
base=graph.Base('G'),
aligned_to=None,
block_id=bid(0)),
graph.Node(node_id=nid(11),
base=graph.Base('T'),
aligned_to=None,
block_id=bid(0))
]
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, 5, 6, 11])])],
graph.SequenceMetadata({'group': '1'})),
msa.SequenceID('seq2'):
graph.Sequence(
msa.SequenceID('seq2'),
[graph.SeqPath([*map(nid, [4, 5, 7, 8, 9, 10, 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),
missings.ConstBaseProvider(self.missing_n), self.metadatacsv)
self.assertEqual(expected_poagraph, actual_poagraph)
def test_1_messy_sequences(self):
maf_path = self.maf_files_dir.joinpath("test_1_messy_sequences.maf")
expected_nodes = [
graph.Node(node_id=nid(0),
base=graph.Base('A'),
aligned_to=None,
block_id=bid(0)),
graph.Node(node_id=nid(1),
base=graph.Base('A'),
aligned_to=nid(2),
block_id=bid(0)),
graph.Node(node_id=nid(2),
base=graph.Base('C'),
aligned_to=nid(1),
block_id=bid(0)),
graph.Node(node_id=nid(3),
base=graph.Base('T'),
aligned_to=None,
block_id=bid(0)),
graph.Node(node_id=nid(4),
base=graph.Base('C'),
aligned_to=nid(5),
block_id=bid(0)),
graph.Node(node_id=nid(5),
base=graph.Base('G'),
aligned_to=nid(4),
block_id=bid(0)),
graph.Node(node_id=nid(6),
base=graph.Base('A'),
aligned_to=None,
block_id=bid(1)),
graph.Node(node_id=nid(7),
base=graph.Base('C'),
aligned_to=None,
block_id=bid(1)),
graph.Node(node_id=nid(8),
base=graph.Base('G'),
aligned_to=None,
block_id=bid(1)),
graph.Node(node_id=nid(9),
base=graph.Base('C'),
aligned_to=nid(10),
block_id=bid(2)),
graph.Node(node_id=nid(10),
base=graph.Base('G'),
aligned_to=nid(9),
block_id=bid(2)),
graph.Node(node_id=nid(11),
base=graph.Base('T'),
aligned_to=None,
block_id=bid(2)),
graph.Node(node_id=nid(12),
base=graph.Base('C'),
aligned_to=None,
block_id=bid(2)),
graph.Node(node_id=nid(13),
base=graph.Base('C'),
aligned_to=None,
block_id=bid(2)),
graph.Node(node_id=nid(14),
base=graph.Base('A'),
aligned_to=None,
block_id=bid(2)),
]
expected_sequences = {
msa.SequenceID('seq0'):
graph.Sequence(
msa.SequenceID('seq0'),
[graph.SeqPath([*map(nid, [1, 3, 4, 6, 8, 9, 11, 12])])],
graph.SequenceMetadata({'group': '1'})),
msa.SequenceID('seq1'):
graph.Sequence(
msa.SequenceID('seq1'),
[graph.SeqPath([*map(nid, [2, 3, 4, 10, 11, 12, 13, 14])])],
graph.SequenceMetadata({'group': '1'})),
msa.SequenceID('seq2'):
graph.Sequence(
msa.SequenceID('seq2'),
[graph.SeqPath([*map(nid, [0, 2, 5, 6, 7, 10, 11, 12, 14])])],
graph.SequenceMetadata({'group': '2'})),
msa.SequenceID('seq3'):
graph.Sequence(msa.SequenceID('seq3'), [],
graph.SequenceMetadata({'group': '2'}))
}
actual_nodes, actual_sequences = maf2poagraph.get_poagraph(
msa.Maf(pathtools.get_file_content_stringio(maf_path), maf_path),
self.metadatacsv)
self.assertEqual(expected_nodes, actual_nodes)
self.assertEqual(expected_sequences, actual_sequences)
