This script is used to conver a "gene" UCSC file into BED12 file

ucscGene2bed12.py

#!/usr/bin/env python
"""
A script that converts a UCSC gene format into BED12 format

Author: Tabea Kischka

Example for UCSC gene format:
594	NM_058167	chr1	-	1253911	1273854	1255202	1267992	7	1253911,1256044,1256991,1257207,1263345,1267861,1273665,	1255487,1256125,1257130,1257310,1263386,1267992,1273854,	0UBE2J2	cmpl	cmpl	0,0,2,1,2,0,-1,
"""
import sys
import csv
import operator

def longblob2list(longblob):
	"""
	Converts a longblob variable into a list of integers.
	"90930917,90931703,90932054," > [90930917,90931703,90932054]
	"""
	if not longblob: return []
	return [int(number) for number in longblob.split(',') if number != "" and number != ","]

ucsc_table_fieldnames = ["bin", "name", "chrom", "strand",
								 "txStart", "txEnd", "cdsStart", "cdsEnd",
								 "exonCount", "exonStarts",
								 "exonEnds", "score", "name2",
								 "cdsStartStat", "cdsEndStat", "exonFrames"]


csvfile = open(sys.argv[1], 'r')
reader = csv.DictReader(csvfile, fieldnames=ucsc_table_fieldnames, dialect='excel-tab')

for row_nr, row in enumerate(reader):
	row['exonEnds'] = longblob2list(row['exonEnds'])
	row['exonStarts'] = longblob2list(row['exonStarts'])
	row['txStart'] = int(row['txStart'])
	p = map(str, [row['chrom'],
		  row['txStart'],
		  row['txEnd'],
		  row['name'],
		  0,
		  row['strand'],
		  row['cdsStart'],
		  row['cdsEnd'],
		  "255,0,255",
		  row['exonCount'],
		  ",".join(map(str, map(operator.sub, row['exonEnds'], row['exonStarts']))),
		  ",".join(map(str, map(operator.sub, row['exonStarts'], [row['txStart']] * len(row['exonStarts']))))
		  ])
	print("\t".join(p))

csvfile.close()