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_poagraph_paths_and_nodes(po_lines: List[str],
sequences_info: Dict[int, POSequenceInfo],
sequences: Dict[msa.SequenceID, graph.Sequence]) -> \
Tuple[List[graph.Node], Dict[msa.SequenceID, graph.Sequence]]:
nodes_count = int(_extract_line_value(po_lines[3]))
paths_count = int(_extract_line_value(po_lines[4]))
nodes: List[graph.Node] = [None] * nodes_count
node_id = 0
for i in range(5 + paths_count * 2, 5 + paths_count * 2 + nodes_count):
node_line = po_lines[i]
base = graph.Base(node_line[0].upper())
in_nodes, po_sequences_ids, aligned_to = _extract_node_parameters(node_line)
sequences_ids = [sequences_info[po_sequences_id].name
for po_sequences_id in po_sequences_ids]
nodes[node_id] = graph.Node(graph.NodeID(node_id),
base,
graph.NodeID(aligned_to))
for seq_id in sequences_ids:
if len(sequences[seq_id].paths) == 1:
sequences[seq_id].paths[0].append(graph.NodeID(node_id))
else:
sequences[seq_id].paths.append(graph.SeqPath([graph.NodeID(node_id)]))
node_id += 1
return nodes, sequences
def nid(x):
return graph.NodeID(x)
def _get_max_node_id(nodes: List[graph.Node]) -> graph.NodeID:
return graph.NodeID(len(nodes) - 1)
def nid(x): return graph.NodeID(x) def bid(x): return graph.BlockID(x)
def nid(x): return graph.NodeID(x) def bid(x): return graph.Base(x)
