def insert_orthology(kobasrc, kobasdir):
organismdb = dbutils.KOBASDB(kobasrc['kobasdb'] + 'organism.db')
for abbr in organismdb.organisms(name=False):
#for abbr in [('tru', ), ('xma', )]:
orthologs = tsv.parse(kobasrc, kobasdir + '/orthology/', abbr[0])
speciesdb = dbutils.KOBASDB(kobasrc['kobasdb'] + abbr[0] + '.db')
speciesdb.con.executemany('INSERT INTO Orthologs VALUES (?, ?)',
orthologs)
def parse(stid_handle, map_handle, kobasrc):
pathways, gene_pathways = {}, {}
stids = {}
for line in stid_handle:
uniprotkb_ac, paid, pname = line.split('\t')[:3]
if not stids.has_key((uniprotkb_ac, pname)):
stids[(uniprotkb_ac, pname)] = paid
for line in map_handle:
uniprotkb_ac, tmp1, pnames, tmp2, sname = line[:-1].split('\t')
if sname not in reactome_organisms.reactome_organisms.keys():
# if sname not in ['H**o sapiens', 'Mus musculus']:
continue
abbr = reactome_organisms.reactome_organisms[sname]
speciesdb = dbutils.KOBASDB(kobasrc['kobasdb'] + abbr + '.db')
gids = speciesdb.gids_from_uniprotkb_ac(uniprotkb_ac)
pnames = re.sub('\[.*\]: ', '', pnames).split('; ')
for gid in gids:
for pname in pnames:
pname = re.sub(' IEA$', '', pname)
if stids.has_key((uniprotkb_ac, pname)):
paid = stids[(uniprotkb_ac, pname)]
else:
paid = ''
if not pathways.setdefault(abbr, {}).has_key(pname):
pathways[abbr][pname] = [0, paid]
gene_pathways.setdefault(abbr, set()).add((gid[0], pname))
for abbr in pathways.keys():
speciesdb = dbutils.KOBASDB(kobasrc['kobasdb'] + abbr + '.db')
count = 0
for pname in pathways[abbr].keys():
count += 1
pathways[abbr][pname][0] = count * 10 + 4
if __name__ == '__main__':
print pathways[abbr][pname][0], 'R', pathways[abbr][pname][
1], pname
else:
speciesdb.con.execute('INSERT INTO Pathways VALUES (?, ?, ?, ?)', \
(pathways[abbr][pname][0], 'R', pathways[abbr][pname][1], pname))
for gene_pathway in gene_pathways[abbr]:
if __name__ == '__main__':
print gene_pathway[0], pathways[abbr][gene_pathway[1]][0]
else:
speciesdb.con.execute('INSERT INTO GenePathways VALUES (?, ?)', \
(gene_pathway[0], pathways[abbr][gene_pathway[1]][0]))
def insert_go(kobasrc, kobasdir):
go_terms, go_inferreds = obo.parse(open(kobasdir + '/go/gene_ontology.1_2.obo'))
for organism in go_organisms.go_organisms:
#for organism in (('gene_association.goa_human', 'not_need', '9606', 'hsa'), \
# ('gene_association.mgi', 'gp2protein.mgi', '10090', 'mmu')):
multi_uniprotkb_ac_gos = association.parse(kobasdir, organism)
abbrs = [abbr for abbr in organism[3].split('|')]
for abbr in abbrs:
speciesdb = dbutils.KOBASDB(kobasrc['kobasdb'] + abbr + '.db')
gos, gene_gos = set(), set()
for uniprotkb_ac_go in multi_uniprotkb_ac_gos[abbr]:
if (uniprotkb_ac_go[1] not in go_categories) and (uniprotkb_ac_go[1] in go_terms.keys()):
gids = speciesdb.gids_from_uniprotkb_ac(uniprotkb_ac_go[0])
for gid in gids:
gene_gos.add((gid[0], uniprotkb_ac_go[1]))
gos.add((uniprotkb_ac_go[1], go_terms[uniprotkb_ac_go[1]]))
for goid in go_inferreds[uniprotkb_ac_go[1]]:
if (goid not in go_categories) and (goid in go_terms.keys()):
gene_gos.add((gid[0], goid))
gos.add((goid, go_terms[goid]))
speciesdb.con.executemany('INSERT INTO Gos VALUES (?, ?)', list(gos))
speciesdb.con.executemany('INSERT INTO GeneGos VALUES (?, ?)', list(gene_gos))
def parse(handle, kobasrc):
pathways, gene_pathways = {}, {}
for line in handle:
paid, pname, tmp1, tmp2, gene = line.split('\t')[:5]
sname = gene.split('|')[0]
if sname not in panther_organisms.panther_organisms.keys():
# if sname not in ['HUMAN', 'MOUSE']:
continue
abbr = panther_organisms.panther_organisms[sname]
speciesdb = dbutils.KOBASDB(kobasrc['kobasdb'] + abbr + '.db')
if 'UniProtKB' in gene:
uniprotkb_ac = gene[gene.find('UniProtKB'):].split('|')[0].split('=')[1]
gids = speciesdb.gids_from_uniprotkb_ac(uniprotkb_ac)
elif 'ENTREZ' in gene:
entrez_gene_id = gene[gene.find('ENTREZ'):].split('|')[0].split('=')[1]
gids = speciesdb.gids_from_entrez_gene_id(entrez_gene_id)
else:
continue
for gid in gids:
if not pathways.setdefault(abbr, {}).has_key(pname):
pathways[abbr][pname] = [0, paid]
gene_pathways.setdefault(abbr, set()).add((gid[0], pname))
for abbr in pathways.keys():
speciesdb = dbutils.KOBASDB(kobasrc['kobasdb'] + abbr + '.db')
count = 0
for pname in pathways[abbr].keys():
count += 1
pathways[abbr][pname][0] = count * 10 + 6
if __name__ == '__main__':
print pathways[abbr][pname][0], 'p', pathways[abbr][pname][1], pname
else:
speciesdb.con.execute('INSERT INTO Pathways VALUES (?, ?, ?, ?)', \
(pathways[abbr][pname][0], 'p', pathways[abbr][pname][1], pname))
for gene_pathway in gene_pathways[abbr]:
if __name__ == '__main__':
print gene_pathway[0], pathways[abbr][gene_pathway[1]][0]
else:
speciesdb.con.execute('INSERT INTO GenePathways VALUES (?, ?)', \
(gene_pathway[0], pathways[abbr][gene_pathway[1]][0]))
def parse(kobasrc, orthologydir, abbr):
orthologs = set()
speciesdb = dbutils.KOBASDB(kobasrc['kobasdb'] + abbr + '.db')
gids = [gid[0] for gid in speciesdb.genes(name=False)]
geid_oeids = get_geid_oeids(orthologydir, abbr)
for gid in gids:
geids = speciesdb.ensembl_gene_ids_from_gid(gid)
for geid in geids:
if geid_oeids.has_key(geid[0]):
for oeid, ospecies in geid_oeids[geid[0]]:
ospeciedb = dbutils.KOBASDB(kobasrc['kobasdb'] + ospecies +
'.db')
oids = ospeciedb.gids_from_ensembl_gene_id(oeid)
for oid in oids:
orthologs.add((gid, oid[0]))
return list(orthologs)
def distr_from_default(default, db, speciesdb, abbr, small_num, idmapping):
distr = {}
if abbr == 'ko':
koids = speciesdb.kos()
for koid in koids:
terms = speciesdb.pathways_from_koid(koid['koid'])
gids = speciesdb.gids_from_koid(koid['koid'], default)
for gid in gids:
if idmapping:
bspeciesdb = dbutils.KOBASDB(config.getrc()['kobasdb'] + default + '.db')
dblink_ids = [dblink_id[0] for dblink_id in bspeciesdb.dblink_ids_from_gid(gid['gid'], idmapping)]
for term in terms:
if idmapping:
distr[(term['name'], db, term['pid'])] = distr.get((term['name'], db, term['pid']), 0) + len(dblink_ids)
else:
distr[(term['name'], db, term['pid'])] = distr.get((term['name'], db, term['pid']), 0) + 1
else:
gids = speciesdb.genes()
for gid in gids:
if db in dbutils.P.keys():
terms = speciesdb.pathways_from_gid(gid['gid'], db)
elif db in dbutils.D.keys():
terms = speciesdb.diseases_from_gid(gid['gid'], db)
elif db in dbutils.G.keys():
terms = speciesdb.gos_from_gid(gid['gid'])
if idmapping:
dblink_ids = [dblink_id[0] for dblink_id in speciesdb.dblink_ids_from_gid(gid['gid'], idmapping)]
if db in dbutils.P.keys() + dbutils.D.keys():
for term in terms:
if idmapping:
distr[(term['name'], db, term['id'])] = distr.get((term['name'], db, term['id']), 0) + len(dblink_ids)
else:
distr[(term['name'], db, term['id'])] = distr.get((term['name'], db, term['id']), 0) + 1
elif db in dbutils.G.keys():
for term in terms:
if idmapping:
distr[(term['name'], db, term['goid'])] = distr.get((term['name'], db, term['goid']), 0) + len(dblink_ids)
else:
distr[(term['name'], db, term['goid'])] = distr.get((term['name'], db, term['goid']), 0) + 1
remove_small_terms2(distr, small_num)
return distr
def read_gmt_db(gene_list, species, dbtype, idtype, min_size, max_size):
gset_name = [] #name of gene sets
gset_des = [] #description of gene sets
hit_genes = [
] #genes in gene sets none of use, only use genes both in list and gene set
D = {
'o': 'OMIM',
'k': 'KEGG DISEASE',
'f': 'FunDO'
} #don't support GAD and N, because lack of did
try:
kobasrc = config.getrc()
speciesdb = dbutils.KOBASDB(kobasrc['kobasdb'] + species + '.db')
dbcatg = dbtype.split(':')[0]
if not dbcatg in ['P', 'D', 'G']:
raise exception.GmtdatabaseErr, "Error Message: This database category you input is not allowed. Please choose 'P','D','G' for database category."
records = []
gset_num = 0
if dbcatg == 'P':
input_dbs = set(dbtype.split(':')[1].split('/'))
avail_dbs = {}
for database in speciesdb.pathwaydbs_from_abbr(species):
avail_dbs[database[0]] = database[1]
dbs = input_dbs.intersection(set(avail_dbs.keys()))
if dbs:
print "Databases: %s" % ', '.join(
[avail_dbs[db] for db in dbs])
else:
raise exception.GmtdatabaseErr, "No supported databases are selected. Supported databases are %s, but your input databases are: %s." % \
('/'.join(avail_dbs.keys()), dbtype)
for db in dbs:
tmp = speciesdb.allpathways(db).fetchall()
num = speciesdb.pathwaynums(db)
records += tmp
gset_num += num
if gset_num == 0:
raise exception.GmtdatabaseErr, "Error Message: \nFail to get information from gene set database."
i = 0
hit_matrix = np.array([[0.0 for s in range(len(gene_list))]
for m in range(gset_num)])
for record in records:
hgene = []
gset_name.append(record[1]) #id
gset_des.append(record[2]) #name
gset_gids = speciesdb.genes_from_pid(
record[0]) #from table GenePathways
for gid in gset_gids:
#print gid
#dblink_id = speciesdb.dblink_ids_from_gid(gid[0], idtype).fetchone()[0]
for each in speciesdb.dblink_ids_from_gid(gid[0], idtype):
dblink_id = each[0]
if dblink_id in gene_list:
hgene.append(dblink_id)
j = gene_list.index(dblink_id)
hit_matrix[i, j] = 1
hit_genes.append(hgene)
i += 1
if not 1 in hit_matrix:
raise exception.GmtdatabaseErr, "Error Message:\nNone of genes in input file(gct file) was matched with genes in gene sets. \nPlease check the gct file(-e), the species(-s), the idtype(-i), the database type(-d)."
elif dbcatg == 'D':
input_dbs = set(dbtype.split(':')[1].split('/'))
avail_dbs = {}
if species != 'hsa':
raise exception.GmtdatabaseErr, "Error Message: Disease is only supported for H**o Sapiens(-s hsa), not supported for this species. Please choose another database category, e.g. 'P','G'"
dbs = input_dbs.intersection(set(D.keys()))
# if dbs:
# print 'Databases: %s' % ', '.join([avail_dbs[db] for db in dbs])
# else:
if not dbs:
raise exception.GmtdatabaseErr, 'No supported databases are selected. Supported databases are %s, but your input databases are: %s.' % \
('/'.join(avail_dbs.keys()), dbtype)
for db in dbs:
tmp = speciesdb.alldiseases(db).fetchall()
num = speciesdb.diseasenums(db)
records += tmp
gset_num += num
if gset_num == 0:
raise exception.GmtdatabaseErr, "Error Message: \nFail to get information from gene set database."
i = 0
hit_matrix = np.array([[0.0 for s in range(len(gene_list))]
for m in range(gset_num)])
for record in records:
hgene = []
#gset_name.append(record[1]) #id
if record[1]:
gset_name.append(record[1])
else:
#gset_name.append(record[2])
continue
gset_des.append(record[2]) #name
gset_gids = speciesdb.genes_from_did(
record[0]) #from table GenePathways
for gid in gset_gids:
#dblink_id = speciesdb.dblink_ids_from_gid(gid[0], idtype).fetchone()[0]
for each in speciesdb.dblink_ids_from_gid(gid[0], idtype):
dblink_id = each[0]
if dblink_id in gene_list:
hgene.append(dblink_id)
j = gene_list.index(dblink_id)
hit_matrix[i, j] = 1
hit_genes.append(hgene)
i += 1
if not 1 in hit_matrix:
raise exception.GmtdatabaseErr, "Error Message:\nNone of genes in input file(gct file) was matched with genes in gene sets. \nPlease check the gct file(-e), the species(-s), the idtype(-i), the database type(-d)."
else:
records = speciesdb.allgoterms().fetchall()
gset_num = speciesdb.gotermnums()
i = 0
hit_matrix = np.array([[0.0 for s in range(len(gene_list))]
for m in range(gset_num)])
for record in records:
hgene = []
gset_name.append(record[0]) #id
gset_des.append(record[1]) #name
gset_gids = speciesdb.genes_from_goid(
record[0]) #from table GenePathways
for gid in gset_gids:
#dblink_id = speciesdb.dblink_ids_from_gid(gid[0], idtype).fetchone()[0]
dblink_id = []
for each in speciesdb.dblink_ids_from_gid(gid[0], idtype):
dblink_id += each[0]
if dblink_id in gene_list:
hgene.append(dblink_id)
j = gene_list.index(dblink_id)
hit_matrix[i, j] = 1
hit_genes.append(hgene)
i += 1
if not 1 in hit_matrix:
raise exception.GmtdatabaseErr, "Error Message:\nNone of genes in input file(gct file) was matched with genes in gene sets. \nPlease check the gct file(-e), the species(-s), the idtype(-i), the database type(-d)."
gset_name = np.array(tuple(gset_name))
gset_des = np.array(gset_des)
hit_genes = np.array(hit_genes)
hitsum = hit_matrix.sum(1)
delindex = np.where((hitsum < min_size) | (hitsum > max_size))[0]
hit_matrix_filtered = np.delete(hit_matrix, delindex, axis=0)
hit_genes_filterd = np.delete(hit_genes, delindex, axis=0)
hit_sum_filtered = np.delete(hitsum, delindex, axis=0)
gset_name_filtered = np.delete(np.array(gset_name), delindex, axis=0)
gset_des_filtered = np.delete(gset_des, delindex, axis=0)
if len(hit_matrix_filtered) == 0 and len(
gset_name_filtered) == 0 and len(
gset_des_filtered) == 0 and len(hit_genes_filterd) == 0:
raise exception.GmtvalueErr, "Error Message:\nAll gene sets" + repr(
len(hit_matrix)
) + "have been filtered. \nPlease check the threshold and ceil of gene set size (values of min_size and max_size). "
except ValueError, e:
sys.exit(e)
PROGRAMS = {'fasta:pro': 'blastp', 'fasta:nuc': 'blastx'}
DBLINKS = {
'id:ncbigene': 'entrez_gene_id',
'id:ncbigi': 'gi',
'id:uniprot': 'uniprotkb_ac',
'id:ensembl': 'ensembl_gene_id'
}
opt_parser, opt, args = config_option()
# KOBAS environment configuration
kobasrc = config.getrc()
# open KOBASDB
organismdb = dbutils.KOBASDB(kobasrc['kobasdb'] + 'organism.db')
if opt.species:
speciesdb = dbutils.KOBASDB(kobasrc['kobasdb'] + opt.species + '.db')
if opt.list:
if opt.species:
print 'Available databases for %s:' % opt.species
databases = speciesdb.databases_from_abbr(opt.species)
for database in databases:
print '\t'.join(database)
else:
print 'Available species: \nko\tKEGG Orthology'
species = organismdb.organisms(name=True)
for specie in species:
print '\t'.join(specie)
sys.exit()
kobasrc['blastdb'] = opt.kobas_home + '/seq_pep/'
if opt.blast_home:
kobasrc['blast_home'] = opt.blast_home
kobasrc['blastp'] = opt.blast_home + '/blastp/'
kobasrc['blastx'] = opt.blast_home + '/blastx/'
if opt.blastdb:
kobasrc['blastdb'] = opt.blastdb + '/'
if opt.kobasdb:
kobasrc['kobasdb'] = opt.kobasdb + '/'
if opt.blastp:
kobasrc['blastp'] = opt.blastp + '/'
if opt.blastx:
kobasrc['blastx'] = opt.blastx + '/'
##open KOBASDB
speciesdb = dbutils.KOBASDB(kobasrc['kobasdb'] + abbr + '.db')
##process opt.db
input_dbs = set(opt.db.split('/'))
avail_dbs = {}
for database in speciesdb.databases_from_abbr(abbr):
avail_dbs[database[0]] = database[1]
dbs = input_dbs.intersection(set(avail_dbs.keys()))
if dbs:
print '##Databases: %s' % ', '.join([avail_dbs[db] for db in dbs])
else:
opt_parser.error('No supported databases are selected. Supported databases are %s, but your input databases are: %s.' % \
('/'.join(avail_dbs.keys()), opt.db))
odistr = discover.Distr()
for pathway in tuple(pathways):
if pathway[0] not in pids:
pathways.remove(pathway)
pathways = tuple(pathways)
gene_pathways = tuple(gene_pathways)
return pathways, gene_pathways
if __name__ == '__main__':
import sys
from pprint import pprint
from kobas import config, dbutils
pathways, gene_pathways = parse(open(sys.argv[1]), \
dbutils.KOBASDB(config.getrc()['kobasdb'] + 'hsa.db'))
print len(pathways), len(gene_pathways)
g, p = set(), set()
for gp in gene_pathways:
g.add(gp[0])
p.add(gp[1])
print len(g), len(p)
pprint(pathways[:5])
pprint(gene_pathways[:5])
gids = bcids[gene_id]
for gid in gids:
gene_pathways.add((gid, pid))
pathways = list(pathways)
gene_pathways = list(gene_pathways)
return pathways, gene_pathways
if __name__ == '__main__':
import sys
from pprint import pprint
from kobas import config, dbutils
pathways, gene_pathways = parse(open(sys.argv[1]), open(sys.argv[2]), \
dbutils.KOBASDB(config.getrc()['kobasdb'] + sys.argv[3] + '.db'))
print len(pathways), len(gene_pathways)
g, p = set(), set()
for gp in gene_pathways:
g.add(gp[0])
p.add(gp[1])
print len(g), len(p)
pprint(pathways[:5])
pprint(gene_pathways[:5])
#!/usr/bin/env python
import os, time, math
import numpy as np
#import rpy2.robjects as robjects
from itertools import combinations
from kobas import discover, annot, dbutils, config
FILENAME = """/rd1/user/tangkj/trunk/test/result_I"""
kobasrc = config.getrc()
speciesdb = dbutils.KOBASDB(kobasrc['kobasdb'] + 'hsa' + '.db')
COLUMN = ('Term', 'Database', 'ID', 'Input number', 'Background number',
'P-Value', 'Corrected P-Value', 'Input', 'Hyperlink')
#robjects.r.library('epiR')
def verify_file(file_name):
##verify file existence and accession, and return file handle
if os.access(file_name, os.F_OK):
return open(file_name)
else:
print 'file %s does not exist' % file_name
sys.exit(1)
def oneTerm(line):
cont = line.split('\t')
term = dict(zip(COLUMN, cont))
return term
