help="The location of the PsimiSQL class")
parser.add_argument('--map-file', required=True, metavar="MapFile", type=str,
help="The location of the file that maps salmonella proteins to uniprot")
parser.add_argument('--dictionary-db', required=True, metavar="Dictionary-db", type=str)
parser.add_argument('--prediction-name', required=True, metavar="PredName", type=str)
args = parser.parse_args()
dictionary_api = sqlite3.connect(args.dictionary_db)
dictionary_cursor = dictionary_api.cursor()
sys.path.append(args.psimisql)
from sqlite_db_api import PsimiSQL
# initiating the memory db
db_api = PsimiSQL()
# the unique nodes will be held here
salmonella_nodes= {}
human_nodes = {}
edges = []
# parsing the map file
with open(args.map_file) as mapFile:
# skipping header
mapFile.readline()
for line in mapFile:
uniprot, salmonella = line.strip().split("\t")
def main():
# opening the old_db for mapping
old_db = PsimiSQL()
old_db.import_from_db_file(SOURCE_DB_LOCATION)
# making the script more verbose
counter = 0
old_db.cursor.execute("SELECT count(*) FROM node")
number_of_nodes = old_db.cursor.fetchone()[0]
# iterating through the old_db's nodes
old_db.cursor.execute("SELECT * FROM node")
# mapping old node_ids to new node old ids
old_node_ids_dict = {}
# initiating an empty db the maped nodes are put
new_db = PsimiSQL()
# declaring a counter to count the nodes that does not match
no_match_counter = 0
invalid_node_counter = 0
# looping through the old_db_s nodes
while True:
row = old_db.cursor.fetchone()
# communicating with user
sys.stdout.write("Querying %d. node from dictionary out of %d\r" %
(counter, number_of_nodes))
counter += 1
# until the last row
if row == None:
break
else:
row_id, mitab_name, alt_accession, mitab_tax_id, pathways, aliases, topology = row
tax_id = str(mitab_tax_id.split(':')[1])
name = str(mitab_name.split(':')[1])
old_node_dict = {
"id": row_id,
"name": mitab_name,
"alt_accession": alt_accession,
"tax_id": mitab_tax_id,
"pathways": pathways,
"aliases": aliases,
"topology": topology
}
# if the fetched node is already mapped, just it's copy will be inserted
# if "uniprot" in mitab_name:
# add_uniprot(old_node_dict,old_node_ids_dict,new_db)
# else:
query = """
SELECT DISTINCT foreign_ids.accession, uniprot.accession, uniprot.is_swissprot, uniprot.is_primary
FROM foreign_ids JOIN uniprot ON foreign_ids.uniprot_id = uniprot.id
WHERE foreign_ids.accession = ? AND uniprot.tax_id = ? AND uniprot.is_primary = 1
"""
tup = (name, tax_id)
DICTIONARY_DB_CURSOR.execute(query, tup)
DICTIONARY_DB.commit()
result = DICTIONARY_DB_CURSOR.fetchall()
if len(result) == 0:
# if there is no match in the map for the current node
no_match_counter += 1
else:
# get a list with only the swissprot nodes from the result of the SQL query
swiss_nodes = get_swiss_arr(result)
# getting the trembl nodes arr
trembl_nodes = get_trembl_arr(result)
# getting the new aliases
aliases = get_aliases_string(trembl_nodes)
# best case scenario it's a 1 -> 1 map
if len(swiss_nodes) == 1:
swiss_accession = "uniprot:" + swiss_nodes[0][1]
add_node(old_node_dict, old_node_ids_dict, swiss_accession,
new_db, aliases)
# if it maps to more than one swissprot accession, all swissprot nodes will be added
elif len(swiss_nodes) > 1:
for node in swiss_nodes:
swiss_accession = "uniprot:" + node[1]
add_node(old_node_dict, old_node_ids_dict,
swiss_accession, new_db, aliases)
# adding trembl nodes if the old node does not match any swissprot accession
else:
for node in trembl_nodes:
trembl_accession = "trembl:" + node[1]
add_node(old_node_dict, old_node_ids_dict,
trembl_accession, new_db, aliases)
print("Inserting to %s nodes done" % SOURCE_DB_TYPE)
# setting up counters, to be able to give the user some information of the ongoing process
old_db.cursor.execute("SELECT count(*) FROM edge")
number_of_edges = old_db.cursor.fetchone()[0]
edge_counter = 0
query = "SELECT * from edge"
old_db.cursor.execute(query)
while True:
# informing the user
sys.stdout.write("Parsing edge # %d out of %d\r" %
(edge_counter, number_of_edges))
row = old_db.cursor.fetchone()
if row == None:
break
else:
edge_counter += 1
# deconstructing the row (list)
edge_id, old_interactor_a_node_id, old_interactor_b_node_id, interactor_a_node_name, interactor_b_node_name, interaction_detection_method, first_author, publication_ids, interaction_types, source_db, interaction_identifiers, confidence_scores, layer = row
# since we get the old interactor id's from this query we can simply look up ther new id(s) in the old_node_ids dict
# it both nodes mapped we add them as an edge to the new db
if old_node_ids_dict.has_key(
old_interactor_a_node_id) and old_node_ids_dict.has_key(
old_interactor_b_node_id):
# looping through every new 'A' node
for new_node_id_a in old_node_ids_dict[
old_interactor_a_node_id]:
new_node_a_dict = new_db.get_node_by_id(new_node_id_a)
# looping through every new 'B' node for every new 'A' node and inserting them as an edge
for new_node_id_b in old_node_ids_dict[
old_interactor_b_node_id]:
new_node_b_dict = new_db.get_node_by_id(new_node_id_b)
# generating the new edge dict
new_edge_dict = {
'interactor_a_node_id': new_node_id_a,
'interactor_b_node_id': new_node_id_b,
'interactor_a_node_name': interactor_a_node_name,
'interactor_b_node_name': interactor_b_node_name,
'interaction_detection_method':
interaction_detection_method,
'first_author': first_author,
'publication_ids': publication_ids,
'source_db': "source database:" + SOURCE_DB_TYPE,
'interaction_types': interaction_types,
'interaction_identifiers': interaction_identifiers,
'confidence_scores': confidence_scores,
'layer': layer
}
# inserting the new node
new_db.insert_edge(new_node_a_dict, new_node_b_dict,
new_edge_dict)
else:
# countering the nodes that can't be inserted to the new db because one of their nodes haven't mapped
invalid_node_counter += 1
print("Inserting edges to %s.db finished!" % SOURCE_DB_TYPE)
new_db.save_db_to_file(DESTINATION_DB_LOCATION)
def main():
# declaring the dicts that will hold the data
nodes = {}
collected_edges = {}
merged_edge_counter = 0
not_merged_edge = 0
# the number of pieces (.db files)
sum_files = len(SOURCE_DB_FILE_LIST)
# filling up the nodes dictionary with the data contained in db piece files
for db_file in SOURCE_DB_FILE_LIST:
# executing a query that selects everything (but the node id) from the current SQLite .db file
db = sqlite3.connect(db_file)
cursor = db.cursor()
cursor.execute("SELECT * FROM node WHERE tax_id = 'taxid:9606' OR tax_id = 'taxid:99284'")
# iterating trough the db row by row
while True:
row = cursor.fetchone()
# until the last row
if row == None:
break
# if unique, inserting the node (row) to the nodes dictionary
id, name, alt_accession, tax_id, pathways, aliases, topology = row
node = {
"name": name,
"alt_accession": alt_accession,
"tax_id": tax_id,
"pathways": pathways,
"aliases": aliases,
"topology": topology,
}
if not nodes.has_key(name):
nodes[name] = node
else:
nodes[name] = get_union_of_nodes(nodes[name], node)
# closing the current db
db.close()
# logging out some info
current_file = SOURCE_DB_FILE_LIST.index(db_file)
sys.stdout.write("Building the node dictionary: Processing %d file out of %d\r" % (current_file, sum_files))
# making a memory database and inserting the unique nodes from the nodes dictionary
print("Inserting nodes to database")
parser = PsimiSQL()
for node in nodes:
parser.insert_unique_node(nodes[node])
nodes[node]["id"] = parser.cursor.lastrowid
# looping through the files again to make an edge list
print("Started building edge dict")
file_counter = 1
for db_file in SOURCE_DB_FILE_LIST:
sys.stdout.write("Inserting edges to edge dict from '%s' (%d/%d)\r" % (db_file, file_counter, sum_files))
# executing a query that selects everything (but the node id) from the current SQLite .db file
db = sqlite3.connect(db_file)
cursor = db.cursor()
cursor.execute("SELECT * FROM edge")
while True:
row = cursor.fetchone()
# if there aren't any more nodes break out of the loop
if not row:
break
else:
# deconstructing the row (list)
edge_row_id, old_interactor_a_node_id, old_interactor_b_node_id, interactor_a_node_name, interactor_b_node_name, interaction_detection_method, first_author, publication_ids, interaction_types, source_db, interaction_identifiers, confidence_scores, layer = (
row
)
# because in the nodes dict building process the query only asks for human and salmonella nodes
# we have to make sure that we don't try to insert edges whose
# nodes are in the nodes dict (=does not a node of other organism)
if nodes.has_key(interactor_a_node_name) and nodes.has_key(interactor_b_node_name):
# generating an edge id that will be the key in the edge dict
edge_id = interactor_a_node_name + "@" + interactor_b_node_name
# generating an edge dict, that will be a value for the key in the collected_edges dict
current_edge = {
"interaction_detection_method": interaction_detection_method,
"first_author": first_author,
"publication_ids": publication_ids,
"interaction_types": interaction_types,
"source_db": source_db,
"interaction_identifiers": interaction_identifiers,
"confidence_scores": confidence_scores,
"layer": layer,
}
# if the collected_edges dict does not contain
# this edge_id. the edge is stored in the collected edges
if not collected_edges.has_key(edge_id):
collected_edges[edge_id] = current_edge
else:
# if collected_edges has this id the edge will be merged
collected_edge = collected_edges[edge_id]
# if an edge is already in the dict it will be merged with the current edge
collected_edge["interaction_types"] = merge_strings(
collected_edge["interaction_types"], current_edge["interaction_types"]
)
collected_edge["first_author"] = merge_strings(
collected_edge["first_author"], current_edge["first_author"]
)
collected_edge["source_db"] = merge_strings(
collected_edge["source_db"], current_edge["source_db"]
)
collected_edge["interaction_identifiers"] = merge_strings(
collected_edge["interaction_identifiers"], current_edge["interaction_identifiers"]
)
collected_edge["interaction_detection_method"] = merge_strings(
collected_edge["interaction_detection_method"], current_edge["interaction_detection_method"]
)
collected_edge["confidence_scores"] = merge_strings(
collected_edge["confidence_scores"], current_edge["confidence_scores"]
)
print("Building edge dict done!")
print("Started inserting edges to the db")
# iterating through edges dictionary and inserting nodes to the SQLite db
for collected_edge_id, edge_to_insert in collected_edges.iteritems():
# getting the nodes
node_a, node_b = collected_edge_id.split("@")
node_a_dict = nodes[node_a]
node_b_dict = nodes[node_b]
parser.insert_edge(node_a_dict, node_b_dict, edge_to_insert)
print("Saving db")
parser.save_db_to_file(DESTINATION)
def main():
# parsing the arguments
arguments = parse_arguments()
# importing the PsimiSQL clas
sys.path.append(arguments.sqlite_db_api)
from sqlite_db_api import PsimiSQL
# the nodes and the edges will be stored in a dict
nodes = {}
edges = {}
# Parsing the nodes first and merging pathways and storing this in the nodes dict created above
# (querying node to check whether it has the same pathway and than updating it with sql queries would be slow)
with open(arguments.source_file) as source_file:
# skipping the header line if needed
if arguments.skip_header:
source_file.readline()
# setting up variables for informing the user
num_lines = float(sum([1 for line in source_file]))
line_counter = float(0)
source_file.seek(0)
# looping through the file
for line in source_file:
# infroming the user
line_counter += 1
if line_counter % 50 == 0:
done = (line_counter / num_lines) * 100
sys.stdout.write("Parsing mitab file (%d%%)\r" % (done))
# deconstructing the line
source_acc, target_acc, source_alt_acc, target_alt_acc, source_alias, target_alias, int_det_method, author, pubmed_ids, source_tax_id, target_tax_id, int_type, source_db, confidence, pathway_ids, layer, source_topology, target_topology = line.strip(
).split("\t")
source_dict = {
"name": source_acc,
"alt_accession": source_alt_acc,
"tax_id": source_tax_id,
"pathways": pathway_ids,
"aliases": source_alias,
"topology": source_topology
}
add_to_nodes(source_dict, nodes)
target_dict = {
"name": target_acc,
"alt_accession": target_alt_acc,
"tax_id": target_tax_id,
"pathways": pathway_ids,
"aliases": target_alias,
"topology": target_topology
}
add_to_nodes(target_dict, nodes)
# adding the edge to the edges dict
edges["%s@%s" % (source_acc, target_acc)] = {
'interaction_detection_method': int_det_method,
'first_author': author,
'publication_ids': pubmed_ids,
'interaction_types': int_type,
'source_db': source_db,
'interaction_identifiers': '-',
'confidence_scores': confidence,
'layer': layer
}
# informing the user
print("Parsing MiTAB file: Finished")
# now that we have the unique nodes we can add them to the Psi-Mi-SQL database
# initiating the memory Mitab database
db_api = PsimiSQL()
num_nodes = float(len(nodes))
line_counter = float(1)
# inserting the nodes to the memory db
for node_name, node_dict in nodes.items():
#informing the user
if line_counter % 50 == 0:
done = float((line_counter / num_nodes) * 100)
sys.stdout.write("Inserting nodes to NetMiTabSQL (%s%%)" % (done))
line_counter += 1
# inserting node to the db file
db_api.insert_node(node_dict)
# updating (mutating) the node dict with the SQL row id in the nodes dictionary so it can be used later
# (again, it is faster to store the row ids fot the rows than querying each rowid)
nodes["id"] = db_api.last_row_id
print("Inserting nodes to NetMiTabSQL: Done")
num_edges = float(len(nodes))
line_counter = float(1)
# inserting the edges to the memory db
for edge_id, edge_dict in edges.items():
#informing the user
if line_counter % 50 == 0:
done = float((line_counter / num_edges) * 100)
sys.stdout.write("Inserting nodes to NetMiTabSQL (%s%%)" % (done))
line_counter += 1
source_name, target_name = edge_id.split('@')
source_dict = nodes[source_name]
target_dict = nodes[target_name]
db_api.insert_edge(source_dict, target_dict, edge_dict)
print("Inserting edges to NetMiTabSQL: Done")
print("Saving the database to filesystem")
# the database is finished, saving
db_api.save_db_to_file(arguments.output_file)
print("Database saved")
}
if not nodes.has_key(name):
nodes[name] = node
else:
nodes[name].update(node)
#closing the current db
db.close()
#logging out some info
current_file = PIECE_LIST.index(filename)
sys.stdout.write(
"Building the node dictionary: Processing %d files out of %d\r" %
(current_file, sum_files))
#making a memory database and inserting the unique nodes from the nodes dictionary
print('Inserting nodes to database')
parser = PsimiSQL(SQL_SEED)
for node in nodes:
parser.insert_unique_node(nodes[node])
#now that we have the nodes in the final db, the edges can be inserted
#there is no need for a edges dictionary, because reading it from the files costs less memory
#iterating through the .db piece files again
print('Inserting edges to database')
for filename in PIECE_LIST:
db = sqlite3.connect(filename)
query = "SELECT * FROM edge"
cursor = db.cursor()
cursor.execute(query)
#iterating trough the current piece .db files
while True:
help="The name and optionally location where the data should be saved.",
)
parser.add_argument(
"--psimisql", required=True, metavar="PsimiSQLLocation", type=str, help="The location of the PsimiSQL class"
)
args = parser.parse_args()
# imports
import sys
sys.path.append(args.psimisql)
from sqlite_db_api import PsimiSQL
# initiating the memory db
db_api = PsimiSQL()
# the unique nodes will be held here
nodes = {}
edges = {}
# parsing the file
with open(args.source_file) as source_file:
# read the first line
source_file.readline()
for line in source_file:
parser.add_argument('--psimisql',
required=True,
metavar="PsimiSQLLocation",
type=str,
help="The location of the PsimiSQL class")
args = parser.parse_args()
# imports
import sys
sys.path.append(args.psimisql)
from sqlite_db_api import PsimiSQL
# initiating the memory db
db_api = PsimiSQL()
# the unique nodes will be held here
nodes = {}
# parsing the file
with open(args.source_file) as source_file:
# skipping the header line
source_file.readline()
for line in source_file:
linearr = line.strip().split("\t")
def main():
# declaring the dicts that will hold the data
nodes = {}
collected_edges = {}
merged_edge_counter = 0
not_merged_edge = 0
# the number of pieces (.db files)
sum_files = len(SOURCE_DB_FILE_LIST)
# filling up the nodes dictionary with the data contained in db piece files
for db_file in SOURCE_DB_FILE_LIST:
# executing a query that selects everything (but the node id) from the current SQLite .db file
db = sqlite3.connect(db_file)
cursor = db.cursor()
cursor.execute(
"SELECT * FROM node WHERE tax_id = 'taxid:9606' OR tax_id = 'taxid:99284'"
)
# iterating trough the db row by row
while True:
row = cursor.fetchone()
# until the last row
if row == None:
break
# if unique, inserting the node (row) to the nodes dictionary
id, name, alt_accession, tax_id, pathways, aliases, topology = row
node = {
"name": name,
'alt_accession': alt_accession,
'tax_id': tax_id,
'pathways': pathways,
'aliases': aliases,
'topology': topology
}
if not nodes.has_key(name):
nodes[name] = node
else:
nodes[name] = get_union_of_nodes(nodes[name], node)
# closing the current db
db.close()
# logging out some info
current_file = SOURCE_DB_FILE_LIST.index(db_file)
sys.stdout.write(
"Building the node dictionary: Processing %d file out of %d\r" %
(current_file, sum_files))
# making a memory database and inserting the unique nodes from the nodes dictionary
print('Inserting nodes to database')
parser = PsimiSQL()
for node in nodes:
parser.insert_unique_node(nodes[node])
nodes[node]['id'] = parser.cursor.lastrowid
# looping through the files again to make an edge list
print("Started building edge dict")
file_counter = 1
for db_file in SOURCE_DB_FILE_LIST:
sys.stdout.write("Inserting edges to edge dict from '%s' (%d/%d)\r" %
(db_file, file_counter, sum_files))
# executing a query that selects everything (but the node id) from the current SQLite .db file
db = sqlite3.connect(db_file)
cursor = db.cursor()
cursor.execute("SELECT * FROM edge")
while True:
row = cursor.fetchone()
# if there aren't any more nodes break out of the loop
if not row:
break
else:
# deconstructing the row (list)
edge_row_id, old_interactor_a_node_id, old_interactor_b_node_id, interactor_a_node_name, interactor_b_node_name, interaction_detection_method, first_author, publication_ids, interaction_types, source_db, interaction_identifiers, confidence_scores, layer = row
# because in the nodes dict building process the query only asks for human and salmonella nodes
# we have to make sure that we don't try to insert edges whose
# nodes are in the nodes dict (=does not a node of other organism)
if nodes.has_key(interactor_a_node_name) and nodes.has_key(
interactor_b_node_name):
# generating an edge id that will be the key in the edge dict
edge_id = interactor_a_node_name + "@" + interactor_b_node_name
# generating an edge dict, that will be a value for the key in the collected_edges dict
current_edge = {
'interaction_detection_method':
interaction_detection_method,
'first_author': first_author,
'publication_ids': publication_ids,
'interaction_types': interaction_types,
'source_db': source_db,
'interaction_identifiers': interaction_identifiers,
'confidence_scores': confidence_scores,
'layer': layer
}
# if the collected_edges dict does not contain
# this edge_id. the edge is stored in the collected edges
if not collected_edges.has_key(edge_id):
collected_edges[edge_id] = current_edge
else:
# if collected_edges has this id the edge will be merged
collected_edge = collected_edges[edge_id]
# if an edge is already in the dict it will be merged with the current edge
collected_edge['interaction_types'] = merge_strings(
collected_edge['interaction_types'],
current_edge['interaction_types'])
collected_edge['first_author'] = merge_strings(
collected_edge['first_author'],
current_edge['first_author'])
collected_edge['source_db'] = merge_strings(
collected_edge['source_db'],
current_edge['source_db'])
collected_edge[
'interaction_identifiers'] = merge_strings(
collected_edge['interaction_identifiers'],
current_edge['interaction_identifiers'])
collected_edge[
'interaction_detection_method'] = merge_strings(
collected_edge['interaction_detection_method'],
current_edge['interaction_detection_method'])
collected_edge['confidence_scores'] = merge_strings(
collected_edge['confidence_scores'],
current_edge['confidence_scores'])
print("Building edge dict done!")
print("Started inserting edges to the db")
# iterating through edges dictionary and inserting nodes to the SQLite db
for collected_edge_id, edge_to_insert in collected_edges.iteritems():
# getting the nodes
node_a, node_b = collected_edge_id.split('@')
node_a_dict = nodes[node_a]
node_b_dict = nodes[node_b]
parser.insert_edge(node_a_dict, node_b_dict, edge_to_insert)
print("Saving db")
parser.save_db_to_file(DESTINATION)
def main():
# opening the old_db for mapping
old_db = PsimiSQL()
old_db.import_from_db_file(SOURCE_DB_LOCATION)
# making the script more verbose
counter = 0
old_db.cursor.execute("SELECT count(*) FROM node")
number_of_nodes = old_db.cursor.fetchone()[0]
# iterating through the old_db's nodes
old_db.cursor.execute("SELECT * FROM node")
# mapping old node_ids to new node old ids
old_node_ids_dict = {}
# initiating an empty db the maped nodes are put
new_db = PsimiSQL()
# declaring a counter to count the nodes that does not match
no_match_counter = 0
invalid_node_counter = 0
# looping through the old_db_s nodes
while True:
row = old_db.cursor.fetchone()
# communicating with user
sys.stdout.write("Querying %d. node from dictionary out of %d\r" % (counter, number_of_nodes))
counter += 1
# until the last row
if row == None:
break
else:
row_id, mitab_name, alt_accession, mitab_tax_id, pathways, aliases, topology = row
tax_id = str(mitab_tax_id.split(':')[1])
name = str(mitab_name.split(':')[1])
old_node_dict = {
"id" : row_id,
"name" : mitab_name,
"alt_accession" : alt_accession,
"tax_id" : mitab_tax_id,
"pathways" : pathways,
"aliases" : aliases,
"topology" : topology
}
# if the fetched node is already mapped, just it's copy will be inserted
# if "uniprot" in mitab_name:
# add_uniprot(old_node_dict,old_node_ids_dict,new_db)
# else:
query = """
SELECT DISTINCT foreign_ids.accession, uniprot.accession, uniprot.is_swissprot, uniprot.is_primary
FROM foreign_ids JOIN uniprot ON foreign_ids.uniprot_id = uniprot.id
WHERE foreign_ids.accession = ? AND uniprot.tax_id = ? AND uniprot.is_primary = 1
"""
tup = (name, tax_id)
DICTIONARY_DB_CURSOR.execute(query, tup)
DICTIONARY_DB.commit()
result = DICTIONARY_DB_CURSOR.fetchall()
if len(result) == 0:
# if there is no match in the map for the current node
no_match_counter+=1
else:
# get a list with only the swissprot nodes from the result of the SQL query
swiss_nodes = get_swiss_arr(result)
# getting the trembl nodes arr
trembl_nodes = get_trembl_arr(result)
# getting the new aliases
aliases = get_aliases_string(trembl_nodes)
# best case scenario it's a 1 -> 1 map
if len(swiss_nodes) == 1:
swiss_accession = "uniprot:"+swiss_nodes[0][1]
add_node(old_node_dict, old_node_ids_dict, swiss_accession, new_db, aliases)
# if it maps to more than one swissprot accession, all swissprot nodes will be added
elif len(swiss_nodes) > 1:
for node in swiss_nodes:
swiss_accession = "uniprot:"+node[1]
add_node(old_node_dict, old_node_ids_dict, swiss_accession, new_db, aliases)
# adding trembl nodes if the old node does not match any swissprot accession
else:
for node in trembl_nodes:
trembl_accession = "trembl:"+node[1]
add_node(old_node_dict, old_node_ids_dict, trembl_accession, new_db, aliases)
print("Inserting to %s nodes done" % SOURCE_DB_TYPE)
# setting up counters, to be able to give the user some information of the ongoing process
old_db.cursor.execute("SELECT count(*) FROM edge")
number_of_edges = old_db.cursor.fetchone()[0]
edge_counter = 0
query = "SELECT * from edge"
old_db.cursor.execute(query)
while True:
# informing the user
sys.stdout.write("Parsing edge # %d out of %d\r" % (edge_counter, number_of_edges))
row = old_db.cursor.fetchone()
if row == None:
break
else:
edge_counter += 1
# deconstructing the row (list)
edge_id, old_interactor_a_node_id, old_interactor_b_node_id, interactor_a_node_name, interactor_b_node_name, interaction_detection_method , first_author, publication_ids, interaction_types, source_db, interaction_identifiers, confidence_scores, layer = row
# since we get the old interactor id's from this query we can simply look up ther new id(s) in the old_node_ids dict
# it both nodes mapped we add them as an edge to the new db
if old_node_ids_dict.has_key(old_interactor_a_node_id) and old_node_ids_dict.has_key( old_interactor_b_node_id):
# looping through every new 'A' node
for new_node_id_a in old_node_ids_dict[old_interactor_a_node_id]:
new_node_a_dict = new_db.get_node_by_id(new_node_id_a)
# looping through every new 'B' node for every new 'A' node and inserting them as an edge
for new_node_id_b in old_node_ids_dict[old_interactor_b_node_id]:
new_node_b_dict = new_db.get_node_by_id(new_node_id_b)
# generating the new edge dict
new_edge_dict = {
'interactor_a_node_id' : new_node_id_a,
'interactor_b_node_id': new_node_id_b,
'interactor_a_node_name' : interactor_a_node_name,
'interactor_b_node_name': interactor_b_node_name,
'interaction_detection_method' : interaction_detection_method,
'first_author' : first_author,
'publication_ids' : publication_ids,
'source_db' : "source database:"+SOURCE_DB_TYPE,
'interaction_types' : interaction_types,
'interaction_identifiers' : interaction_identifiers,
'confidence_scores' : confidence_scores,
'layer' : layer
}
# inserting the new node
new_db.insert_edge(new_node_a_dict, new_node_b_dict, new_edge_dict)
else:
# countering the nodes that can't be inserted to the new db because one of their nodes haven't mapped
invalid_node_counter += 1
print("Inserting edges to %s.db finished!" % SOURCE_DB_TYPE)
new_db.save_db_to_file(DESTINATION_DB_LOCATION)
help="The location of the source file for nodes")
parser.add_argument('--outfile', required=True, metavar="OutputFile", type=str,
help="The name and optionally location where the data should be saved.")
parser.add_argument('--psimisql', required=True, metavar="PsimiSQLLocation", type=str,
help="The location of the PsimiSQL class")
args = parser.parse_args()
# imports
import sys
sys.path.append(args.psimisql)
from sqlite_db_api import PsimiSQL
# initiating the memory db
db_api = PsimiSQL()
# the unique nodes will be held here
nodes = {}
edges = {}
# parsing the file
with open(args.source_file) as source_file:
# read the first line
source_file.readline()
for line in source_file:
metavar="Dictionary-db",
type=str)
parser.add_argument('--prediction-name',
required=True,
metavar="PredName",
type=str)
args = parser.parse_args()
dictionary_api = sqlite3.connect(args.dictionary_db)
dictionary_cursor = dictionary_api.cursor()
sys.path.append(args.psimisql)
from sqlite_db_api import PsimiSQL
# initiating the memory db
db_api = PsimiSQL()
# the unique nodes will be held here
salmonella_nodes = {}
human_nodes = {}
edges = []
# parsing the map file
with open(args.map_file) as mapFile:
# skipping header
mapFile.readline()
for line in mapFile:
uniprot, salmonella = line.strip().split("\t")
help="The name and optionally location where the data should be saved.")
parser.add_argument('--db-name', required=True, metavar="DatabaseName", type=str,
help="The name of the source database")
parser.add_argument('--psimisql', required=True, metavar="PsimiSQLLocation", type=str,
help="The location of the PsimiSQL class")
args = parser.parse_args()
# imports
import sys
sys.path.append(args.psimisql)
from sqlite_db_api import PsimiSQL
# initiating the memory db
db_api = PsimiSQL()
# parsing nodes
# nodes will be stored temporarily in memory, because re-querying them is slower
nodes = {}
# looping through the node file
with open(args.node_source_file) as node_file:
# informing the user
print "Parsing nodes"
sum_nodes = sum([1 for x in node_file])
progress = 1
node_file.seek(0)
help="The location of the csv that was saved from arn")
parser.add_argument('--outfile', required=True, metavar="OutputFile", type=str,
help="The name and optionally location where the data should be saved.")
parser.add_argument('--psimisql', required=True, metavar="PsimiSQLLocation", type=str,
help="The location of the PsimiSQL class")
args = parser.parse_args()
# imports
import sys
sys.path.append(args.psimisql)
from sqlite_db_api import PsimiSQL
# initiating the memory db
db_api = PsimiSQL()
# extracting nodes
nodes = {}
edges = {}
print("Parsing file")
with open(args.source_file) as arn_file:
# skipping the first line
arn_file.readline()
for line in arn_file:
# splitting up the line
help="The name and optionally location where the data should be saved.",
)
parser.add_argument(
"--psimisql", required=True, metavar="PsimiSQLLocation", type=str, help="The location of the PsimiSQL class"
)
args = parser.parse_args()
# imports
import sys
sys.path.append(args.psimisql)
from sqlite_db_api import PsimiSQL
# initiating the memory db
db_api = PsimiSQL()
# the unique nodes will be held here
nodes = {}
# parsing the file
with open(args.source_file) as source_file:
# skipping the header line
source_file.readline()
for line in source_file:
linearr = line.strip().split("\t")
def main():
for db in args.source_files:
# informing the user
print("Parsing %s" % db)
cursor = sqlite3.connect(db).cursor()
mapped_nodes = {}
nodemap = {}
cursor.execute("SELECT * FROM node")
result = cursor.fetchall()
length = len(result)
current = 1
new_db = PsimiSQL()
cursor.execute("SELECT count(*) FROM node")
num_of_nodes = cursor.fetchone()[0]
# mapping nodes
print("Mapping nodes")
for line in result:
# informing user
if (current % 50 == 0):
print("Mapping nodes %d/%d" % (current, length))
current += 1
row_id, name, alt_accession, tax_id, pathways, aliases, topology = line
old_uniprot = name
new_uniprot = "uniprot:"+get_primary(old_uniprot.split(':')[1])
# storing the new uniprot id for every old id
nodemap[old_uniprot] = new_uniprot
mapped_node = {
'name': new_uniprot,
'alt_accession': alt_accession,
'tax_id': tax_id,
'pathways': pathways,
'aliases': aliases,
'topology': topology
}
mapped_node['id'] = new_db.insert_node(mapped_node)
mapped_nodes[new_uniprot] = mapped_node
if len(nodemap) != num_of_nodes:
print "Gebasz"
# mapping edges
cursor.execute("SELECT * FROM edge")
result = cursor.fetchall()
print("Mapping edges")
length = len(result)
current = 1
shit_counter = 0
for row in result:
if (current % 10 == 0):
print("Parsing edge %d/%d" % (current, length))
current += 1
old_source_uniprot = row[3]
old_target_uniprot = row[4]
edge_dict = {
'interaction_detection_method': row[5],
'first_author': row[6],
'publication_ids': row[7],
'interaction_types': row[8],
'source_db': row[9],
'interaction_identifiers': row[10],
'confidence_scores': row[11],
'layer': "0"
}
if (old_source_uniprot not in mapped_nodes or old_target_uniprot not in mapped_nodes):
shit_counter +=1
else:
new_db.insert_edge(mapped_nodes[old_source_uniprot], mapped_nodes[old_target_uniprot], edge_dict)
# saving the mapped db and informing user
db_name = os.path.split(db)[1]
print("Saving db to %s " % (args.outdir+"/mapped"+db_name))
print("SHITCOUNTER %d" % shit_counter )
new_db.save_db_to_file(args.outdir+"/mapped"+db_name)
