Mercurial > repos > yating-l > regtools_junctions_extract

comparison validator.py @ 0:01ed8e112f2a draft

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
planemo upload commit 6bf6d4ec8ff2ec6bd45b5f483ff2f83b6229d57d
author yating-l
date Wed, 12 Apr 2017 17:44:58 -0400
parents
children
comparison
equal deleted inserted replaced
-1:000000000000 0:01ed8e112f2a
1 import sys
2
3 """
4 Call checkAndFixBed, check the integrity of bed file. If the strand is not "+" or "-" truncate that line and report to users
5 create a column and move the score column to that column.
6 """
7 def checkAndFixBed(bedfile, revised_file):
8 # Store the lines that have been removed
9 removedLines = []
10 # Remove the lines with invalid strand, create a score column to store the original scores and change scores in the original score column all to 1000
11 with open(revised_file, 'w') as tmp:
12 with open(bedfile, 'r') as f:
13 lines = f.readlines()
14 i = 1
15 for line in lines:
16 fields = line.split()
17 strand = fields[5]
18 score = fields[4]
19 fields[4] = '1000'
20 fields.append(score)
21 if (strand == '+' or strand == '-'):
22 tmp.write('\t'.join(map(str, fields)))
23 tmp.write("\n")
24 else:
25 removedLines.append("line" + str(i) + ": " + line)
26 i = i+1
27
28 return removedLines
29
30 def main():
31 inputfile = str(sys.argv[1])
32 outputfile = str(sys.argv[2])
33 removed = checkAndFixBed(inputfile, outputfile)
34 if (removed != []):
35 print "\nRemoved invalid lines: \n"
36 print "\n".join(removed)
37
38 if __name__ == "__main__":
39 main()