StacksVCFMod.py

#This script modifies a VCF outputted by Stacks 1.19 to the correct BP position
#This version of Stacks outputs a physical position that is + 1 for positive
#strand tags and - 1 for negative strand tags.
#The script should be run from a directory with a .csv file exported from the
#MySQL catalog_tags table. The csv file should contain which 2 columns: 
#Column 1 is a list of Catalog Tag IDs
#Column 2 is a '+' or '-'
#The script automatically detects a csv file in the working directory and uses it
#The use is prompted for the name of the vcf file to be modified.

import vcf, os, csv, glob, sys
filename = glob.glob('*.csv')

#read in csv with info on + or - strand information, two columns tag_id, strand
csvfile = open(filename[0], "rb")
csvreader = csv.reader(csvfile)
StrandDict = dict()
for row in csvreader:
        StrandDict[str(row[0])] = row[1]

#read in your vcf file
vcfname = raw_input("Enter the full name of the vcf to be modified:\n")
print "Thank you, now modifying your vcf and saving to " + vcfname.split(".")[0] + "-mod.vcf ..."
vcf_reader = vcf.Reader(open(vcfname,'r'))
PosDict = dict()
for record in vcf_reader:
    SNPID = str(record.ID) + '_' + str(record.POS)
    PosDict[SNPID] = record.POS

for key in PosDict:
    tag_id = key.split("_")[0]
    if StrandDict[tag_id] == '+':
        PosDict[key] += 1
    else:
        PosDict[key] -= 1

vcf_reader2= vcf.Reader(open(vcfname,'r'))
modVcfname = vcfname.split(".")[0] + "-mod.vcf"
vcf_writer = vcf.Writer(open(modVcfname, 'w'), vcf_reader2)
for record in vcf_reader2:
    tag_id = str(record.ID) + '_' + str(record.POS)
    record.POS = PosDict[tag_id]
    vcf_writer.write_record(record)
    
print "done!"