CATNAPparsedvcf

#This will parse out the SNPs from chr 7 CNTNAP2 gene into a vcf file, from phase 3 1000 genomes project, all subjects genome build 37
#Jennifer Owen
#Stephanie Perez Robles
#Lisa Dimusto
import sys

infile = 'ALL.chr7.phase3_shapeit2_mvncall_integrated_v5a.20130502.genotypes.vcf'

class ParseV:

    def __init__(self,infile,List=None,samplestuff=None):
        self.infile = infile
        self.List = []
        self.samplestuff=[] # creates a new empty list, think this addresses the problem from last time where the mistaken use of class variable with a mutable data type?
        with open ('ALL.chr7.phase3_shapeit2_mvncall_integrated_v5a.20130502.genotypes.vcf','r') as file:
            for line in file.readlines():
                line=line.rstrip()
                if line.startswith('##'):
                    continue
                elif line.startswith('#'):
                    parts= line.split('\t')
                    self.samplestuff= parts[:]
                    #print(samplestuff)
                else:
                    if line.startswith('7'):
                        P=line.split()
                        POS=P[1]
                        if '146116035' <= POS <= '148420998':
                            ID = P[2]
                            CHROM = P[0]
                            POSITION= P[1]
                            Ref = P[3]
                            ALT = P[4]
                            QUAL= P[5]
                            FILTER= P[6]
                            INFO= P[7]
                            FORMAT= P[8]
                            FIRST=P[9:]
                            self.List.append([CHROM,ID,POSITION,Ref,ALT,QUAL,FILTER,INFO,FORMAT,FIRST])
                            #print(self.List)
#vcf.close # close file
#return List


    def outfile(self): # need to work on this part
        out = open('chr7CATNAP.vcf','w')
        out.write(str(self.samplestuff) +'\n'+ str(self.List))
        #for i in self.List:
        #    out.write(str(i[0]) + "\t" + str(i[1]) + "\t" + str(i[2]) + "\t" + str(i[3]) + "\t" + str(i[4]) + "\t" + str(i[5]) + "\t" + str(i[6]) + "\t" + str(i[7]) + "\t" + str(i[8]) + "\n")
        out.close # close file

tired = ParseV(infile)
tired.outfile()