def convert_all_children_coords(shortest_path_nodes, children, parent):
shortest_path_nodes.sort(key=lambda x: x.query_block_start)
mapped_children = {}
total_bases, mismatches, insertions, deletions = 0, 0, 0, 0
for child in children:
total_bases += (child.end - child.start + 1)
nearest_start_coord, nearest_end_coord = find_nearest_aligned_start_and_end(
child.start, child.end, shortest_path_nodes, parent)
if nearest_start_coord != -1 and nearest_end_coord != -1:
lifted_start, start_node = convert_coord(nearest_start_coord,
shortest_path_nodes)
lifted_end, end_node = convert_coord(nearest_end_coord,
shortest_path_nodes)
deletions += find_deletions(start_node, end_node,
shortest_path_nodes)
deletions += (nearest_start_coord -
child.start) + (child.end - nearest_end_coord)
mismatches += find_mismatched_bases(child.start, child.end,
shortest_path_nodes, parent)
insertions += find_insertions(start_node, end_node,
shortest_path_nodes)
strand = get_strand(shortest_path_nodes[0], parent)
new_child = new_feature.new_feature(
child.id, child.featuretype,
shortest_path_nodes[0].reference_name, 'Liftoff', strand,
min(lifted_start, lifted_end) + 1,
max(lifted_start, lifted_end) + 1, dict(child.attributes))
mapped_children[new_child.id] = new_child
else:
deletions += (child.end - child.start + 1)
return mapped_children, (total_bases - deletions) / total_bases, (total_bases - mismatches - (insertions +
deletions)) / \
total_bases
def add_children(parent_dict, child_dict, lowest_children, feature_db):
c = feature_db.conn.cursor()
cond = ', '.join('"{0}"'.format(w) for w in lowest_children)
query = "select * from relations join features on features.id = relations.child where relations.child IN ({})".format(
cond)
c.execute(query)
results = c.fetchall()
added_children_ids = []
for result in results:
feature_tup = tuple(result)
parent = feature_tup[0]
if parent in parent_dict:
child = new_feature.new_feature(feature_tup[3], feature_tup[6],
feature_tup[4], feature_tup[5],
feature_tup[10], feature_tup[7],
feature_tup[8],
json.loads(feature_tup[12]))
if child.featuretype != "intron":
if "Parent" not in child.attributes:
add_parent_tag(child, feature_db)
child_dict[parent].append(child)
added_children_ids.append(child.id)
single_level_features = np.setdiff1d(lowest_children, added_children_ids)
for feature in single_level_features:
if feature in parent_dict:
child_dict[feature] = [parent_dict[feature]]
def make_new_parent(feature_list, parent, feature_hierarchy):
children = [feature for feature in feature_list.values() if
"Parent" in feature.attributes and feature.attributes["Parent"][0] == parent]
starts, ends = [child.start for child in children], [child.end for child in children]
ref_parent = get_ref_parent(parent, feature_hierarchy)
target_parent_feature = new_feature.new_feature(ref_parent.id, ref_parent.featuretype, children[0].seqid,
'Liftoff', children[0].strand, min(starts), max(ends),
dict(ref_parent.attributes))
feature_list[target_parent_feature.id] = target_parent_feature
return target_parent_feature
def add_intermediates(intermediate_ids, intermediate_dict, feature_db):
c = feature_db.conn.cursor()
cond = ', '.join('"{0}"'.format(w) for w in intermediate_ids)
query = "select * from features where id IN ({})".format(cond)
for result in c.execute(query):
feature_tup = tuple(result)
intermediate_feature = new_feature.new_feature(
feature_tup[0], feature_tup[3], feature_tup[1], feature_tup[2],
feature_tup[7], feature_tup[4], feature_tup[5],
json.loads(feature_tup[9]))
intermediate_dict[intermediate_feature.id] = intermediate_feature
if "Parent" not in intermediate_feature.attributes:
add_parent_tag(intermediate_feature, feature_db)
def add_parents(parent_dict, child_dict, highest_parents, parent_types_to_lift,
feature_db):
c = feature_db.conn.cursor()
cond = ', '.join('"{0}"'.format(w) for w in highest_parents)
query = "SELECT * FROM features WHERE id IN ({})".format(cond)
for result in c.execute(query):
feature_tup = tuple(result)
parent = new_feature.new_feature(feature_tup[0], feature_tup[3],
feature_tup[1], feature_tup[2],
feature_tup[7], feature_tup[4],
feature_tup[5],
json.loads(feature_tup[9]))
if parent.featuretype in parent_types_to_lift:
parent_dict[parent.id] = parent
child_dict[parent.id] = []
