imputation_cleaining

impute2cleaning_narcolepsyGWAS_Apr22.py

import glob
import pandas as pd
import matplotlib.pyplot as plt
import numpy as np
import re
def cleanimpute(x, y, z):
	cleaned_info = x[(x['info'] >= 0.8) & (x['certainty'] >= 0.8)]#x[x['certainty'] >= 0.9]
	cleaned_impute = y[y[1].isin(cleaned_info.rs_id)]
	cleaned_impute.to_csv(z, index=False, header=None)

import re
p =re.compile(r'[.]')
def todos(x, y):
	n=0
	for line in x:
		line1 = line.replace("\n", "")
		fields = line1.split(",")
		if len(p.findall(fields[1])) is 0:
			n += 1
			snp = fields[1]
			pos = fields[2]
			chr = fields[0]
			A1 = fields[3]
			A2 = fields[4]
			out = [chr, snp, pos, A1, A2]
			fields1 = np.array(fields[5:], dtype = float)
			for i in range(0, len(fields1), 3):
				array1 = fields1[i:i+3]
				array2 = round(array1[0]*0+array1[1]*1+array1[2]*2, 2)
				out.append(str(array2))
		#print out
			y.write(" ".join(out) + "\n")
	print "##################################processed", n, "SNPS from", x, "dosages written to", y
	y.close()
	x.close()

######################impute2 cleaning ###################################
for s in range(1, 23):
	pathtoimpute= '/media/labcomp/HDD3/GWAS/CHR' + str(s) + "_pEUR.impute2"
	pathtoinfo =  '/media/labcomp/HDD/GWASProjects/NarcolepsyGWASdata/imputeextras/chr' + str(s) + ".plate_EUR.info"
	x = pd.read_table(pathtoinfo, delim_whitespace= True, header = 0)
	y= pd.read_table(pathtoimpute, delim_whitespace= True, header = None)
	z= "/media/labcomp/HDD3/GWAS/fil_April22_pEUR_CHR" + str(s) + ".impute2"
	cleanimpute(x, y, z)

###########################convert to dosage#############################
for s in range(1, 23):
	pathtoimpute = '/media/labcomp/HDD3/GWAS/fil_April22_pEUR_CHR' + str(s) + '.impute2'
	pathtoout ='/media/labcomp/HDD3/GWAS/pEUR_April22_CHR' + str(s) + '.dos'
	x = open(pathtoimpute, 'r')
	y= open(pathtoout, 'w')
	todos(x, y)

################################################
################################################

for s in range(1, 23):
	pathtoimpute= '/media/labcomp/HDD/GWASProjects/NarcolepsyGWASdata/CHR' + str(s) + "_P35_36_55.impute2"
	pathtoinfo =  '/media/labcomp/HDD/GWASProjects/NarcolepsyGWASdata/imputeextras/chr' + str(s) + ".Plate35_36_55.info"
	x = pd.read_table(pathtoinfo, delim_whitespace= True, header = 0)
	y= pd.read_table(pathtoimpute, delim_whitespace= True, header = None)
	y[0] = s
	z= "/media/labcomp/HDD3/GWAS/fil_April22_p35_36_55_CHR" + str(s) + ".impute2"
	cleanimpute(x, y, z)

import re
p = re.compile(r'[.]')
for s in range(1, 23):
	pathtoimpute = '/media/labcomp/HDD3/GWAS/fil_April22_p35_36_55_CHR' + str(s) + '.impute2'
	pathtoout ='/media/labcomp/HDD3/GWAS/p35_36_55_April22_CHR' + str(s) + '.dos'
	x = open(pathtoimpute, 'r')
	y= open(pathtoout, 'w')
	n=0
	todos(x, y)

##########################################


for s in range(1, 23):
	pathtoimpute= '/media/labcomp/HDD3/GWAS/EAS_clean_CHR' + str(s) + ".impute2"
	pathtoinfo =  '/media/labcomp/HDD/GWASProjects/NarcolepsyGWASdata/imputeextras/chr' + str(s) + ".plate_EAS.info"
	x = pd.read_table(pathtoinfo, delim_whitespace= True, header = 0)
	y= pd.read_table(pathtoimpute, delim_whitespace= True, header = None)
	z="/media/labcomp/HDD3/GWAS/fil_April22_pEAS_CHR" + str(s) + ".impute2"
	cleanimpute(x, y, z)

import re
p = re.compile(r'[.]')
for s in range(1, 23):
	pathtoimpute = '/media/labcomp/HDD3/GWAS/fil_April22_pEAS_CHR' + str(s) + '.impute2'
	pathtoout ='/media/labcomp/HDD3/GWAS/pEAS_April22_CHR' + str(s) + '.dos'
	x = open(pathtoimpute, 'r')
	y= open(pathtoout, 'w')
	n=0
	todos(x, y)


##########################################
for s in range(1, 23):
	pathtoimpute= '/media/labcomp/HDD3/GWAS/plate86to87_CHR' + str(s) + ".impute2"
	pathtoinfo =  '/media/labcomp/HDD/GWASProjects/NarcolepsyGWASdata/imputeextras/chr' + str(s) + ".p86to87QC.info"
	x = pd.read_table(pathtoinfo, delim_whitespace= True, header = 0)
	y= pd.read_table(pathtoimpute, delim_whitespace= True, header = None)
	z="/media/labcomp/HDD3/GWAS/fil_April22_p86to87_CHR" + str(s) + ".impute2"
	cleanimpute(x, y, z)

import re
p = re.compile(r'[.]')
for s in range(1, 23):
	pathtoimpute = '/media/labcomp/HDD3/GWAS/fil_April22_p86to87_CHR' + str(s) + '.impute2'
	pathtoout ='/media/labcomp/HDD3/GWAS/p86to87_April22_CHR' + str(s) + '.dos'
	x = open(pathtoimpute, 'r')
	y= open(pathtoout, 'w')
	n=0
	todos(x, y)

###########################################Q17
/media/labcomp/HDD/GWASProjects/japanesedata_GWAS/clean_mar10_CHR" + str(s) + "_Q35.impute2

import re
p = re.compile(r'[.]')
for s in range(1, 23):
	pathtoimpute = '/media/labcomp/HDD/GWASProjects/japanesedata_GWAS/clean_mar10_CHR' + str(s) + '_Q35.impute2'
	pathtoout ='/media/labcomp/HDD3/GWAS/Q35_CHR' + str(s) + '.dos'
	x = open(pathtoimpute, 'r')
	y= open(pathtoout, 'w')
	n=0
	todos(x, y)



import re
p = re.compile(r'[.]')
for s in range(1, 23):
	pathtoimpute = '/media/labcomp/HDD/GWASProjects/japanesedata_GWAS/clean_mar10_CHR' + str(s) + '_Q17.impute2'
	pathtoout ='/media/labcomp/HDD3/GWAS/Q17_CHR' + str(s) + '.dos'
	x = open(pathtoimpute, 'r')
	y= open(pathtoout, 'w')
	n=0
	todos(x, y)

############################# write annon functions to process impute2 files ##########################
### old function -see above
def cleanimpute(x, y, z):
	cleaned_info = x[(x['info'] >= 0.3) & (x['certainty'] >= 0.8)]#x[x['certainty'] >= 0.9]
	cleaned_impute = y[y[1].isin(cleaned_info.rs_id)]
	cleaned_impute.to_csv(z, index=False, header=None)





##function to perform impute2 to dosage conversion
x = open('/media/labcomp/HDD3/GWAS/test.impute2', 'r')
y= open('/media/labcomp/HDD3/GWAS/test.dos', 'w')
import re
p =re.compile(r'[.]')
def todos(x, y):
	n=0
	for line in x:
		line1 = line.replace("\n", "")
		fields = line1.split(",")
		if len(p.findall(fields[1])) is 0:
			n += 1
			snp = fields[1]
			pos = fields[2]
			chr = fields[0]
			A1 = fields[3]
			A2 = fields[4]
			out = [chr, snp, pos, A1, A2]
			fields1 = np.array(fields[5:], dtype = float)
			for i in range(0, len(fields1), 3):
				array1 = fields1[i:i+3]
				array2 = round(array1[0]*0+array1[1]*1+array1[2]*2, 2)
				out.append(str(array2))
		#print out
			y.write(" ".join(out) + "\n")
	print "##################################processed", n, "SNPS from", x, "dosages written to", y
	y.close()
	x.close()


import re
p =re.compile(r'[.]')
for s in range(1, 23):
	pathtoimpute = '/media/labcomp/HDD3/GWAS/filter_CHR' + str(s) + '_p86to87.impute2'
	pathtoout ='/media/labcomp/HDD3/GWAS/p86to87_CHR' + str(s) + '.dos'
	x = open(pathtoimpute, 'r')
	y= open(pathtoout, 'w')
	todos(x, y)

################################################# add sample IDS to genotype files ################


pref_cols = ['CHR', 'SNP', 'POS', 'A1', 'A2']
for rows in open('/media/labcomp/HDD3/GWAS/p86to87_order.sample', 'r'):
	line = rows.replace('\n', '')
	line1 = line.split('\t')[1]
	pref_cols.append(line1)
out = open('/media/labcomp/HDD3/GWAS/test86to87header', 'w')
out.write(" ".join(pref_cols))
out.close()

########write an annon function to implement

def addcols(x, y):
	pref_cols = ['CHR', 'SNP', 'POS', 'A1', 'A2']
	for rows in x:
		line = rows.replace('\n', '')
		line1 = line.split('\t')[1]
		pref_cols.append(line1)
	y.write(" ".join(pref_cols)+"\n")
	y.close()
	x.close()
########apply this function over a loop

files = open('/media/labcomp/HDD3/GWAS/filelist_April22', 'r')
for i in files:
	plateID = i.replace('\n', '')
	plateID1 = plateID.split('_')[0]
	print plateID1, plateID
	pathtosample = '/media/labcomp/HDD3/GWAS/' + str(plateID)
	pathtoout = '/media/labcomp/HDD3/GWAS/' + str(plateID1) + "_colnames"
	x= open(pathtosample, 'r')
	y=open(pathtoout, 'w')
	addcols(x, y)

################add sample names in bash paralell ############################## EMonly
parallel --jobs 8 cat pEUR_colnames pEUR_April22_CHR{}.dos '>' final_dos/pEUR_April22_CHR{}.dos ::: {1..22}
parallel --jobs 8 cat p86to87_colnames p86to87_April22_CHR{}.dos '>' final_dos/p86to87_April22_CHR{}.dos ::: {1..22}
parallel --jobs 8 cat p35_colnames p35_36_55_April22_CHR{}.dos '>' final_dos/p35_36_55_April22_CHR{}.dos ::: {1..22}
#parallel --jobs 8 cat p37_colnames p37_CHR{}.dos '>' final_dos/p37_CHR{}.dos ::: {1..22}
#parallel --jobs 8 cat p38_colnames p37_CHR{}.dos '>' final_dos/p38_CHR{}.dos ::: {1..22}
parallel --jobs 8 cat pEAS_colnames pEAS_April22_CHR{}.dos '>' final_dos/pEAS_April22_CHR{}.dos ::: {1..22}
#parallel --jobs 8 cat p77to85_colnames p77to85_CHR{}.dos '>' final_dos/p77to85_CHR{}.dos ::: {1..22}
#parallel --jobs 8 cat Q17_colnames Q17_CHR{}.dos '>' final_dos/Q17_CHR{}.dos ::: {1..22}
#parallel --jobs 8 cat Q35_colnames Q35_CHR{}.dos '>' final_dos/Q35_CHR{}.dos ::: {1..22}

##################join all files ## to make on dos file for each chr
import pandas as pd
from pandas import Series, DataFrame
filelist =[]
for rows in open('/media/labcomp/HDD3/GWAS/final_dos/filelist_CHR1', 'r'):
	filelist.append(rows.replace('\n', ''))

colstodrop=['CHR', 'POS', 'A1', 'A2']

pEUR_CHR1 = pd.read_table(filelist[-3], delim_whitespace=True, header=0, index_col='SNP', low_memory =True)
pEUR_CHR1.drop(colstodrop, axis=1, inplace=True)
p35_36_55_CHR1 = pd.read_table(filelist[0], delim_whitespace=True, header=0, index_col='SNP')
p35_36_55_CHR1.drop(colstodrop, axis=1, inplace=True)
Q17_CHR1 = pd.read_table(filelist[-2], delim_whitespace=True, header=0, index_col='SNP')
Q17_CHR1.drop(colstodrop, axis=1, inplace=True)
Q35_CHR1 = pd.read_table(filelist[-1], delim_whitespace=True, header=0, index_col='SNP')
Q35_CHR1.drop(colstodrop, axis=1, inplace=True)
p86to87_CHR1 = pd.read_table(filelist[4], delim_whitespace=True, header=0, index_col='SNP')
p86to87_CHR1=p86to87_CHR1.join(pEUR_CHR1, how='outer')
p86to87_CHR1=p86to87_CHR1.join(p35_36_55_CHR1, how='outer')
p86to87_CHR1=p86to87_CHR1.join(Q17_CHR1, how='outer')
p86to87_CHR1=p86to87_CHR1.join(Q35_CHR1, how='outer')

for names in filelist:
	fileid = names.split('.')[0]
p86to87_CHR1.drop(['A1', 'A2'], axis=1, inplace=True)

p86to87_CHR1.to_csv('/media/labcomp/HDD3/GWAS/final_dos/CHR1_merged.dos', index=True, header =True)



import pandas as pd
import numpy as np

for s in range(1, 23):
	a=pd.read_csv('/media/labcomp/HDD3/GWAS/final_dos/nodups_p86to87_CHR'+str(s)+'.dos', header=0, index_col='SNP', low_memory =True)
	a['CHR'] = s
	b=pd.read_table('/media/labcomp/HDD3/GWAS/final_dos/pEUR_CHR'+str(s)+'.dos', delim_whitespace=True, header=0, index_col='SNP', low_memory =True)
	c=pd.read_table('/media/labcomp/HDD3/GWAS/final_dos/p35_36_55_CHR'+str(s)+'.dos', delim_whitespace=True, header=0, index_col='SNP', low_memory =True)
	d=pd.read_table('/media/labcomp/HDD3/GWAS/final_dos/Q17_CHR'+str(s)+'.dos', delim_whitespace=True, header=0, index_col='SNP', low_memory =True)
	e=pd.read_table('/media/labcomp/HDD3/GWAS/final_dos/Q35_CHR'+str(s)+'.dos', delim_whitespace=True, header=0, index_col='SNP', low_memory =True)
	f=pd.read_table('/media/labcomp/HDD3/GWAS/final_dos/pEAS_CHR'+str(s)+'.dos', delim_whitespace=True, header=0, index_col='SNP', low_memory =True)
	filename ='/media/labcomp/HDD3/GWAS/final_dos/merged_dupsrm_CHR'+str(s)+'.dos'
	dosage_merge(a, b, c, d, e, f, filename)
	print 'done with chr', str(s), 'written to', filename


def dosage_merge(a, b, c, d, e, f, filename):
	colstodrop = ['CHR', 'POS', 'A1', 'A2']
	b.drop(colstodrop, axis=1, inplace=True)
	b.drop_duplicates(keep='last', inplace=True)
	c.drop(colstodrop, axis=1, inplace=True)
	c.drop_duplicates(keep='last', inplace=True)
	d.drop(colstodrop, axis=1, inplace=True)
	d.drop_duplicates(keep='last', inplace=True)
	e.drop(colstodrop, axis=1, inplace=True)
	e.drop_duplicates(keep='last', inplace=True)
	f.drop(colstodrop, axis=1, inplace=True)
	f.drop_duplicates(keep='last', inplace=True)
	a=a.join(b, how='outer') #pEUR
	a=a.join(c, how='outer') #p35_36_55
	a=a.join(d, how='outer') #Q17
	a=a.join(e, how='outer') #Q35
	a=a.join(f, how='outer') #pEAS
	a.drop(['A1', 'A2'], axis=1, inplace=True)
	a.to_csv(filename, index=True, header=True)



def dosage_merge(a, b, c, d, e, f, filename):
	colstodrop = ['CHR', 'POS', 'A1', 'A2']
	b.drop(colstodrop, axis=1, inplace=True)
	c.drop(colstodrop, axis=1, inplace=True)
	d.drop(colstodrop, axis=1, inplace=True)
	e.drop(colstodrop, axis=1, inplace=True)
	f.drop(colstodrop, axis=1, inplace=True)
	f=f.index.drop_duplicates(keep='last')
	a=a.join(b, how='outer') #pEUR
	a=a.join(c, how='outer') #p35_36_55
	a=a.join(d, how='outer') #Q17
	a=a.join(e, how='outer') #Q35
	a=a.join(f, how='outer') #pEAS
	a.drop(['A1', 'A2'], axis=1, inplace=True)
	a.to_csv(filename, index=True, header=True)

################dos cleaning of p86to87

chr1 = pd.read_table('/media/labcomp/HDD3/GWAS/final_dos/p86to87_CHR1.dos', header=0, delim_whitespace=True)
A1=chr1[chr1['A1'] == '-']
A2=chr1[chr1['A2'] == '-']
A1_A2=pd.concat([A1, A2])

##get all duplicated SNPs
dupchr1=chr1[chr1['SNP'].duplicated(keep=False)]
chr1.drop_duplicates(['SNP'], keep=False, inplace =True)
dupfil=dupchr1[(dupchr1.A1 != '-') & (dupchr1.A2 != '-')]
dupfil2=dupfil[(dupfil.A1 != '0') & (dupfil.A2 != '0')]
total=pd.concat([chr1, dupfil2])
total.drop_duplicates(['SNP'], keep=Last, inplace=True)

##############remove dups and clean 
import numpy as np
import pandas as pd

def filterdups(a, output):
	dupchr = a[a['SNP'].duplicated(keep=False)]
	a.drop_duplicates(['SNP'], keep=False, inplace =True)
	dupfil = dupchr[(dupchr.A1 != '-') & (dupchr.A2 != '-')]
	dupfil2 = dupfil[(dupfil.A1 != '0') & (dupfil.A2 != '0')]
	total=pd.concat([a, dupfil2])
	total.drop_duplicates(['SNP'], keep='last', inplace=True)
	total.to_csv(output, index=False, header=True)

for s in range(1, 23):
	a=pd.read_table('/media/labcomp/HDD3/GWAS/final_dos/p86to87_April22_CHR'+str(s)+'.dos', delim_whitespace=True, header=0, low_memory =True)
	output = '/media/labcomp/HDD3/GWAS/final_dos/nodups_p86to87_April22_CHR'+str(s)+'.dos'
	filterdups(a, output)








#analysis of only EM samples, make doasage for 
#p86to87
import pandas as pd
import numpy as np

for s in range(1, 23):
	a=pd.read_csv('/media/labcomp/HDD3/GWAS/final_dos/nodups_p86to87_April22_CHR'+str(s)+'.dos', header=0, index_col='SNP', low_memory =True)
	a['CHR'] = s
	b=pd.read_table('/media/labcomp/HDD3/GWAS/final_dos/pEUR_April22_CHR'+str(s)+'.dos', delim_whitespace=True, header=0, index_col='SNP', low_memory =True)
	c=pd.read_table('/media/labcomp/HDD3/GWAS/final_dos/p35_36_55_April22_CHR'+str(s)+'.dos', delim_whitespace=True, header=0, index_col='SNP', low_memory =True)
	#d=pd.read_table('/media/labcomp/HDD3/GWAS/final_dos/Q17_CHR'+str(s)+'.dos', delim_whitespace=True, header=0, index_col='SNP', low_memory =True)
	#e=pd.read_table('/media/labcomp/HDD3/GWAS/final_dos/Q35_CHR'+str(s)+'.dos', delim_whitespace=True, header=0, index_col='SNP', low_memory =True)
	d=pd.read_table('/media/labcomp/HDD3/GWAS/final_dos/pEAS_April22_CHR'+str(s)+'.dos', delim_whitespace=True, header=0, index_col='SNP', low_memory =True)
	filename ='/media/labcomp/HDD3/GWAS/final_dos/EMplates_April22_CHR'+str(s)+'.dos'
	dosage_merge(a, b, c, d, filename)
	print 'done with chr', str(s), 'written to', filename



def dosage_merge(a, b, c, d, filename):
	colstodrop = ['CHR', 'POS', 'A1', 'A2']
	b.drop(colstodrop, axis=1, inplace=True)
	b.drop_duplicates(keep='last', inplace=True)
	c.drop(colstodrop, axis=1, inplace=True)
	c.drop_duplicates(keep='last', inplace=True)
	d.drop(colstodrop, axis=1, inplace=True)
	d.drop_duplicates(keep='last', inplace=True)
	a=a.join(b, how='inner') #pEUR
	a=a.join(c, how='inner') #p35_36_55
	a=a.join(d, how='inner') #pEAS
	a.drop(['A1', 'A2'], axis=1, inplace=True)
	a.to_csv(filename, index=True, header=True)
###########################################################################concat all chrs to make one big dosage file################################################
import pandas as pd
doslist = open('EM_April23_doslist', 'r')
dosages=[]
for i in doslist:
	dos=i.replace('\n', '')
	tempdos = pd.read_csv(dos, index_col='SNP', header=0, low_memory=True, verbose =True)
	dosages.append(tempdos)
total_dos = pd.concat(dosages)
### calcualte maf for each snp
import numpy as np
maf =[]
for i in range(0, len(total_dos)):
	tempmean= np.array([np.mean(total_dos.ix[i, 2:])/2, 1-(np.mean(total_dos.ix[i, 2:])/2)])
	maf.append(np.min(tempmean))


###################################################################asscoiation testing for QTLs#########################################################################
a= open('header_full', 'r')
for line in a:
	print line

line=line.replace('\r', '')
line=line.replace('\n', '')
line=line.split('\t')

b = open('full', 'r')
out2=[]
n =0
for line in b:
	n += 1
	if n >1:
		#print line
		line = line.replace('\n', '')
		line = line.replace('\r', '')
		fields = line.split('\t')
		#print fields
		fields2 = []
		for i in fields:
			if len(i) is 0:
				fields2.append('NA')
			else:
				fields2.append(i)
		#print len(fields2)
		out2.append(fields2)

	
final_out=pd.DataFrame(out2)
final_out.columns = line


fullassoc = pd.read_csv('full', header = 0, usecols=line)  

###read in the association files(EM, EM_CSF, EM_SERUm, JP, EM_CS_SERUMRATIO)
EM_CSF= pd.read_csv('TOTAL_EMonly_QTLs.csv', header=0, delim_whitespace=True)
EM_CSF_merge=pd.merge(EM_CSF, final_out, left_on='snps', right_on='SNPS', how='inner')
EM_CSF_merge.sort_values(by='pvalue', ascending=True, inplace=True)
EM_CSF_merge.to_csv('EM_CSF_ASSOCIATION_WITH_DIS_TRAIT.csv', index=False)

EM_SERUM = pd.read_csv('TOTAL_EMonly_Serum_QTLs.csv', header=0)
EM_SERUM_merge=pd.merge(EM_SERUM, final_out, left_on='snps', right_on='SNPS', how='inner')
EM_SERUM_merge.sort_values(by='pvalue', ascending=True, inplace=True)
EM_SERUM_merge.to_csv('EM_SERUM_ASSOCIATION_WITH_DIS_TRAIT.csv', index=False)

EM_ratios = pd.read_csv('TOTAL_ratio_QTLs.csv', header=0)
EM_ratios_merge=pd.merge(EM_ratios, final_out, left_on='snps', right_on='SNPS', how='inner')
EM_ratios_merge.sort_values(by='pvalue', ascending=True, inplace=True)
EM_ratios_merge.to_csv('EM_Ratios_ASSOCIATION_WITH_DIS_TRAIT.csv', index=False)

##########################################################################################
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import re
from pandas import DataFrame

EM_CSF = pd.read_csv('EM_CSF_QTLs_Apr3.csv', 'r')

fields=[]
x = open('Gene_Clusters.txt', 'r')
for line in x:
	line2 = line.replace('\t', ';')
	line3 = line2.split('\n')
	print line3
	fields.append(line3)

for f in fields:
	f2 = "".join(f)
	print f2











###########################get all hardgenotypes
import pandas as pd
cd /media/labcomp/HDD/GWASProjects/NarcolepsyGWASdata/
bimlist = ['HCRT_batch2_AA_3.bim', 'Plate35_36_55.bim', 'EAS_plate.bim', 'EUR_plate.bim']
path = '/media/labcomp/HDD/GWASProjects/NarcolepsyGWASdata/'
bims=[]
for i in bimlist:
	temp=pd.read_csv(path+i, delim_whitespace=True, header=None)
	temp['plateID'] = i
	bims.append(temp)
bims[0].plateID = 'Plate86to87'
EM_genotypedsnps = pd.concat(bims)

EM_genotypedsnps=EM_genotypedsnps.rename(columns={0:'CHR', 1:'RSID', 3:'POS', 4:'A1', 5:'A2'})


#p86to87 =  HCRT_batch2_AA_3.bim
#pEUR = EUR_plate
#p77to85QC = '/media/labcomp/HDD/GWASProjects/NarcolepsyGWASdata/Plate77to85QC_F.bim'
#p35_55_36 = Plate35_36_55
#pEAS = EAS_plate.bim

require(reshape2)
em_genos = fread('~/Desktop/EM_genotyped_snps.csv', header = T)
agg_em=dcast(data = em_genos, formula = CHR+RSID+POS+A1+A2 ~ plateID, fun.aggregate = length)
agg_em2=dcast(data = em_genos, formula = CHR+RSID+POS ~ plateID, fun.aggregate = length)
write.csv(agg_em2, file = '~/Desktop/EM_genotypes_summarized_April22.csv', row.names = F)