def importRepeatsFromUCSC(infile, outfile, ucsc_database, repeattypes, genome):
'''import repeats from a UCSC formatted file.
The repeats are stored as a :term:`gff` formatted file.
'''
repclasses = "','".join(repeattypes.split(","))
# Repeats are either stored in a single ``rmsk`` table (hg19) or in
# individual ``rmsk`` tables (mm9) like chr1_rmsk, chr2_rmsk, ....
# In order to do a single statement, the ucsc mysql database is
# queried for tables that end in rmsk.
dbhandle = PipelineUCSC.connectToUCSC(
host=PARAMS["ucsc_host"],
user=PARAMS["ucsc_user"],
database=ucsc_database)
cc = dbhandle.execute("SHOW TABLES LIKE '%%rmsk'")
tables = [x[0] for x in cc.fetchall()]
if len(tables) == 0:
raise ValueError("could not find any `rmsk` tables")
tmpfile = P.getTempFile(shared=True)
total_repeats = 0
for table in tables:
E.info("%s: loading repeats from %s" % (ucsc_database, table))
cc = dbhandle.execute(
"""SELECT genoName, 'repeat', 'exon', genoStart+1, genoEnd, '.',
strand, '.',
CONCAT('class \\"', repClass, '\\"; family \\"', repFamily, '\\";')
FROM %(table)s
WHERE repClass in ('%(repclasses)s') """ % locals())
n = 0
for data in cc.fetchall():
n += 1
tmpfile.write("\t".join(map(str, data)) + "\n")
E.info("%s: %s=%i repeats downloaded" % (ucsc_database, table, n))
total_repeats += n
if total_repeats == 0:
raise ValueErrror("did not find any repeats for %s" % ucsc_database)
tmpfile.close()
tmpfilename = tmpfile.name
statement = '''cat %(tmpfilename)s
| %(pipeline_scriptsdir)s/gff_sort pos
| cgat gff2gff
--method=sanitize
--sanitize-method=genome
--skip-missing
--genome-file=%(genome)s
--log=%(outfile)s.log
| gzip
> %(outfile)s
'''
P.run()
os.unlink(tmpfilename)
