LiftRsNumber.py

From Genome Analysis Wiki
Revision as of 11:33, 10 August 2011 by Zhanxw (talk | contribs)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigationJump to search
#!/usr/bin/python
import sys, os

def usage():
    print("%s file: lift over rs number. " % sys.argv[0] )
    print("file should look like:")
    print("11111")
    print("11112")
    print("...")
def myopen(fn):
    import gzip
    try:
        h = gzip.open(fn)
        ln = h.read(2) # read arbitrary bytes so check if @param fn is a gzipped file
    except:
        # cannot read in gzip format
        return open(fn)
    h.close()
    return gzip.open(fn)

RS_HISTORY = set() # store rs
RS_MERGE = dict() # high_rs -> (lower_rs, current_rs)
if __name__ == '__main__':
    if len(sys.argv) != 2:
	usage()
	sys.exit(1)
	
    # record obsolete rs number
    for ln in myopen('/net/wonderland/home/zhanxw/mydata/dbSNP/SNPHistory.bcp.gz'):
	fd = ln.strip().split('\t')
	if ln.find('re-activ') < 0:
	    RS_HISTORY.add(fd[0])

    # record rs number merge history
    for ln in myopen('/net/wonderland/home/zhanxw/mydata/dbSNP/RsMergeArch.bcp.gz'):
	fd = ln.strip().split('\t')
	h, l = fd[0], fd[1]
	c = fd[6]
	# print 'c=', c
	RS_MERGE[h] = (l, c)

    # # find some retracted SNPs for fun
    # for h, v in RS_MERGE.iteritems():
    # 	l, c = v
    # 	if c in RS_HISTORY:
    # 	    print h, '->', l

    import re
    rsPattern = re.compile(r'[0-9]*')
    for ln in myopen(sys.argv[1]):
    	rs = ln.strip()
	if not rsPattern.match(rs):
	    print 'ERROR: rs number should be like "1000"'
	    sys.exit(2)
	# rs number not appear in RS_MERGE -> there is no merge on this rs
	if rs not in RS_MERGE:
	    print "unchanged\t", rs
	    continue
	# lift rs number
	while True:
	    if rs in RS_MERGE:
		rsLow, rsCurrent = RS_MERGE[rs]
		if rsCurrent not in RS_HISTORY:
		    print "lifted\t", rsCurrent
		    break
		else:
		    rs = rsLow
	    else:
		print "unlifted\t", rs
		break