int main(int argc, char **argv)
{
SNP **snpmarkers ;
Indiv **indivmarkers ;
int numsnps, numindivs ;
unsigned char *packg1, *packg2 ;
int **snppos ;
int *snpindx ;
int lsnplist, lindlist, numeg ;
int i,j;
SNP *cupt, *cupt1, *cupt2, *cupt3 ;
Indiv *indx ;
int ch1, ch2 ;
int fmnum , lmnum ;
int num, n1, n2 ;
int nkill = 0 ;
int t, k, x ;
int nignore, numrisks = 1 ;
char **genolist ;
int numgenolist ;
int maxmiss ;
tersem = YES ; // no snp counts
readcommands(argc, argv) ;
setomode(&outputmode, omode) ;
packmode = YES ;
settersemode(tersem) ;
nums1 =
getsnps(snp1, &snpm1, 0.0, NULL, &nignore, numrisks) ;
putped(1) ;
freeped() ;
nums2 =
getsnps(snp2, &snpm2, 0.0, NULL, &nignore, numrisks) ;
putped(2) ;
freeped() ;
for (x=0; x<nums1; ++x) {
cupt1 = snpm1[x] ;
cupt1 -> tagnumber = -1 ;
}
for (x=0; x<nums2; ++x) {
cupt2 = snpm2[x] ;
t = x %1000 ;
// if (t==0) printf("zz %d %d\n", x, nums2) ;
k = snpindex(snpm1, nums1, cupt2 -> ID) ;
if (k<0) {
cupt2 -> ignore = YES ;
continue ;
}
cupt1 = snpm1[k] ;
cupt1 -> tagnumber = x ;
t = checkmatch(cupt1, cupt2) ;
if (t==1) continue ;
if (t==2) {
cupt2 -> isrfake = YES ;
continue ;
}
if (t<0) {
cupt1 -> ignore = cupt2 -> ignore = YES ;
continue ;
}
printf("allele funny: %s", cupt1 -> ID) ;
printalleles(cupt1, stdout) ;
printalleles(cupt2, stdout) ;
printnl() ;
cupt1 -> ignore = cupt2 -> ignore = YES ;
continue ;
}
freesnpindex() ;
numi1 = getindivs(ind1, &indm1) ;
numi2 = getindivs(ind2, &indm2) ;
for (x=0; x<numi2; ++x) {
k = indindex(indm1, numi1, indm2[x] -> ID) ;
// this code could be modified to allow duplicate individuals
if (k>=0) fatalx("dup ind: %s\n", indm2[x] -> ID) ; // fix later?
}
setgenotypename(&geno1, ind1) ;
getped(1) ;
getgenos(geno1, snpm1, indm1,
nums1, numi1, nignore) ;
packg1 = (unsigned char *) getpackgenos() ;
clearpackgenos() ;
setgenotypename(&geno2, ind2) ;
getped(2) ;
getgenos(geno2, snpm2, indm2,
nums2, numi2, nignore) ;
packg2 = (unsigned char *) getpackgenos() ;
numindivs = mergeit(snpm1, snpm2, &indm1, indm2, nums1, nums2, numi1, numi2) ;
snpmarkers = snpm1 ;
numsnps = nums1 ;
indivmarkers = indm1 ;
free(packg1) ;
free(packg2) ;
outfiles(snpoutfilename, indoutfilename, genooutfilename,
snpmarkers, indivmarkers, numsnps, numindivs, packout, ogmode) ;
printf("##end of mergeit run\n") ;
return 0 ;
}
int main(int argc, char **argv)
{
int **snppos ;
int *snpindx ;
char **snpnamelist, **indnamelist ;
char **eglist ;
int lsnplist, lindlist, numeg ;
int i,j;
SNP *cupt, *cupt1, *cupt2, *cupt3 ;
Indiv *indx ;
double gpos1,gpos2,cpos1,cpos2,gd, cd, gd100 ;
double rthresh, zt ;
int mpflag, ret, numvalidind, nvalid, numvalidsnps ;
int ch1, ch2 ;
int fmnum , lmnum ;
int num, n1, n2 ;
int nkill = 0 ;
int t, k, g ;
int nindiv = 0, e, f, lag=1 ;
double xc[9], xd[4], xc2[9] ;
double ychi, zscore, zthresh = 20.0 ;
double y1, y2 ;
int nignore, numrisks = 1 ;
char **genolist ;
int numgenolist ;
char c1, c2 ;
int t1, t2 ;
malexhet = YES ; // convertf default is don't change the data
tersem = YES ; // no snp counts
readcommands(argc, argv) ;
setomode(&outputmode, omode) ;
packmode = YES ;
settersemode(tersem) ;
if (r2thresh > 0.0) killr2 = YES ;
if (badpedignore) setbadpedignore() ;
numsnps =
getsnps(snpname, &snpmarkers, 0.0, badsnpname, &nignore, numrisks) ;
for (i=0; i<numsnps; i++) {
if (xchrom == -1) break ;
cupt = snpmarkers[i] ;
if (cupt -> chrom != xchrom) cupt -> ignore = YES ;
if (cupt -> ignore) continue ;
t = nnint(cupt -> physpos) ;
if ( (t< lopos) || (t >hipos)) cupt -> ignore = YES ;
}
nignore = 0 ;
for (i=0; i<numsnps; i++) {
cupt = snpmarkers[i] ;
if (cupt -> chrom > maxchrom) cupt -> ignore = YES ;
if (cupt -> chrom < minchrom) cupt -> ignore = YES ;
if (cupt -> ignore) ++nignore ;
}
if (numsnps == nignore) fatalx("no valid snps\n") ;
numindivs = getindivs(indivname, &indivmarkers) ;
if (polarid != NULL) {
polarindex = indindex(indivmarkers, numindivs, polarid) ;
if (polarindex<0) fatalx("polarid %s not found\n") ;
}
if (genotypelist!= NULL) {
getgenos_list(genotypelist, snpmarkers, indivmarkers,
numsnps, numindivs, nignore) ;
}
else {
setgenotypename(&genotypename, indivname) ;
getgenos(genotypename, snpmarkers, indivmarkers,
numsnps, numindivs, nignore) ;
}
if (newsnpname != NULL) {
nums2 =
getsnps(newsnpname, &snpm2, 0.0, NULL, &nignore, numrisks) ;
remap(snpmarkers, numsnps, snpm2, nums2) ;
snpmarkers = snpm2 ;
numsnps = nums2 ;
}
if (newindivname != NULL) {
numind2 = getindivs(newindivname, &indm2) ;
remapind(snpmarkers, numsnps, indivmarkers, indm2, numindivs, numind2) ;
indivmarkers = indm2 ;
numindivs = numind2 ;
if (polarid != NULL) {
polarindex = indindex(indivmarkers, numindivs, polarid) ;
}
}
if (mkdiploid) {
numindivs = rmindivs(snpmarkers, numsnps, indivmarkers, numindivs) ;
numind2 = mkindh2d(indivmarkers, &indm2, numindivs) ;
remaph2d(snpmarkers, numsnps, indivmarkers, indm2, numindivs, numind2) ;
indivmarkers = indm2 ;
numindivs = numind2 ;
}
if (deletedup) dedupit(snpmarkers, numsnps) ; // only one marker per position
for (i=0; i<numsnps; i++) {
cupt = snpmarkers[i] ;
if (zerodistance) cupt -> genpos = 0.0 ;
c1 = cupt -> alleles[0] ;
c2 = cupt -> alleles[1] ;
t1 = pedval(&c1) % 5 ;
t2 = pedval(&c2) % 5 ; // 0 and 5 are no good
if ((t1==0) && (t2 >0)) flip1(cupt, phasedmode, YES) ;
}
flipstrand(flipstrandname, snpmarkers, numsnps) ;
flipsnps(flipsnpname, snpmarkers, numsnps, phasedmode) ;
if (polarindex>=0) {
for (i=0; i<numsnps; i++) {
cupt = snpmarkers[i] ;
g = getgtypes(cupt, polarindex) ;
if (g==0) {
printf("polarizing %s\n", cupt -> ID) ;
flip1(cupt, NO, YES) ;
g = getgtypes(cupt, polarindex) ;
if (g!=2) fatalx("badbug\n") ;
}
if (g != 2) cupt -> ignore = YES ;
}
}
if (outputall) {
outfiles(snpoutfilename, indoutfilename, genooutfilename,
snpmarkers, indivmarkers, numsnps, numindivs, packout, ogmode) ;
printf("##end of convertf run (outputall mode)\n") ;
return 0 ;
}
if (poplistname != NULL)
{
ZALLOC(eglist, numindivs, char *) ;
numeg = loadlist(eglist, poplistname) ;
seteglist(indivmarkers, numindivs, poplistname);
for (i=0; i<numindivs; ++i) {
indx = indivmarkers[i] ;
if (indx -> affstatus == NO) indx -> ignore = YES ;
}
}
else
setstatus(indivmarkers, numindivs, "Case") ;
numsnps = rmsnps(snpmarkers, numsnps, deletesnpoutname) ;
numindivs = rmindivs(snpmarkers, numsnps, indivmarkers, numindivs) ;
if (killr2) {
nkill = killhir2(snpmarkers, numsnps, numindivs, r2physlim, r2genlim, r2thresh) ;
if (nkill>0) printf("killhir2. number of snps killed: %d\n", nkill) ;
}
if ( nhwfilter > 0 ) {
hwfilter(snpmarkers, numsnps, numindivs, nhwfilter, deletesnpoutname);
}
if ( xregionname ) {
excluderegions(xregionname, snpmarkers, numsnps, deletesnpoutname);
}
numvalidind = 0 ;
for (i=0; i<numindivs; ++i) {
indx = indivmarkers[i] ;
if (indx -> ignore) continue ;
if (numvalidgtind(snpmarkers, numsnps, i) ==0) {
indx -> ignore = YES ;
printf("no data for individual: %s\n", indx -> ID) ;
}
if (indx -> ignore == NO) ++numvalidind ;
}
if (maxmiss<0) maxmiss = (int) (maxmissfrac * (double) numvalidind+1) ;
printf("numvalidind: %5d maxmiss: %5d\n", numvalidind, maxmiss) ;
if (numvalidind == 0) fatalx("no valid samples!\n") ;
for (k=0; k<numsnps; ++k) {
if (maxmiss>numvalidind) break ;
cupt = snpmarkers[k] ;
t = numvalidind - numvalidgtypes(cupt) ;
// printf("zz %20s %4d %4d\n", cupt -> ID, t, numvalidind-t) ;
if (maxmiss < t) {
cupt -> ignore = YES ;
}
/**
if (numvalidind == t) {
printf("no data for snp: %s\n", cupt -> ID) ;
cupt -> ignore = YES ;
}
*/
}
if (fastdup) {
printf("fastdup set %d\n", fastdupnum) ;
if (fastdupnum > 0) {
setfastdupnum(fastdupnum) ;
setfastdupthresh(fastdupthresh, fastdupkill) ;
fastdupcheck(snpmarkers, indivmarkers, numsnps, numindivs) ;
}
}
if (decim>0) {
snpdecimate(snpmarkers, numsnps, decim, dmindis, dmaxdis) ;
}
outfiles(snpoutfilename, indoutfilename, genooutfilename,
snpmarkers, indivmarkers, numsnps, numindivs, packout, ogmode) ;
printf("##end of convertf run\n") ;
return 0 ;
}
int main(int argc, char **argv)
{
double *V;
double *xx;
double *iscase;
double *iscasecorr;
int K;
int k,m,n;
int nignore;
double rowsum, rowsum1;
double chisq, Echisq, gamma, denom;
readcommands(argc, argv) ;
if (outputname != NULL)
openit(outputname, &fpout, "w") ;
else
fpout = stdout;
fprintf(fpout, "Chisq EIGENSTRAT\n");
setinmode(&inmode, imode);
packmode = YES;
numsnps =
getsnps(snpname, &snpmarkers, 0.0, NULL, &nignore, 1) ;
NSAMPLES = getindivs(indivname, &indivmarkers) ;
setstatus(indivmarkers, NSAMPLES, "Case") ;
setgenotypename(&genotypename, indivname) ;
if (genotypename != NULL) {
getgenos(genotypename, snpmarkers, indivmarkers,
numsnps, NSAMPLES, nignore) ;
}
/*******************************************************************/
/* Free memory: Usually this is done in outfiles: */
/* */
/* nind = rmindivs(&snpmarkers, numsnps, &indmarkers, NSAMPLES); */
/* */
/* But where is the snpmarkers array released? */
/*******************************************************************/
L = numpc;
readpcafile(&V, &outlier, &K, L, NSAMPLES);
getphenos(NSAMPLES, &iscase, outlier, &iscasecorr, L, V);
/* main eigenstrat loop here */
if ((xx = (double *)malloc(NSAMPLES*sizeof(*xx))) == NULL)
{ fprintf(stderr,"CM\n"); exit(1); }
for(m=0;m<numsnps;m++) {
SNP *cupt = snpmarkers[m];
for(n=0; n<NSAMPLES; n++)
{
int j = getgtypes(cupt,n);
if(j == 0) { xx[n] = 0.0; }
else if(j == 1) { xx[n] = 0.5; }
else if(j == 2 ) { xx[n] = 1.0; }
else if(j == -1) { xx[n] = -100.0; }
if(outlier[n] == 1) xx[n] = -100.0;
}
/* mean-adjust xx */
rowsum = 0.0; rowsum1 = 0.0;
for(n=0; n<NSAMPLES; n++)
{
if(qtmode == NO && ((outlier[n]) || (xx[n] < -99.0))) continue;
if(qtmode == YES && ((outlier[n]) || (xx[n] == -100.0))) continue;
rowsum += xx[n];
rowsum1 += 1.0;
}
for(n=0; n<NSAMPLES; n++)
{
if(outlier[n]) continue;
if(qtmode == NO) {
if (xx[n] < -99.0)
xx[n] = -100.0; /* still keep track */
else
xx[n] -= rowsum/rowsum1;
}
else {
if (xx[n] == -100.0)
xx[n] = -100.0; /* still keep track */
else
xx[n] -= rowsum/rowsum1;
}
}
/* Chisq */
chisq = compute_chisq(xx,iscase);
/* EIGENSTRAT */
for(k=0; k<L; k++)
{
gamma = 0.0;
denom = 0.0;
for(n=0; n<NSAMPLES; n++)
{
if(qtmode == NO && (outlier[n] || xx[n] < -99.0)) continue;
if(qtmode == YES && (outlier[n] || xx[n] == -100.0)) continue;
gamma += xx[n]*V[NSAMPLES*n+k];
denom += V[NSAMPLES*n+k]*V[NSAMPLES*n+k];
}
gamma /= denom;
for(n=0; n<NSAMPLES; n++)
{
if(qtmode == NO && (outlier[n] || xx[n] < -99.0)) continue;
if(qtmode == YES && (outlier[n] || xx[n] == -100.0)) continue;
xx[n] -= gamma*V[NSAMPLES*n+k];
}
}
Echisq = compute_chisqE(xx,iscasecorr);
if(rowsum1 == 0.0)
{
chisq = -1.0; Echisq = -1.0;
}
if(chisq >= 0.0) fprintf(fpout,"%.04f",chisq);
else fprintf(fpout,"NA");
if(Echisq >= 0.0) fprintf(fpout," %.04f\n",Echisq);
else fprintf(fpout," NA\n");
if(NSAMPLES*m > MAXSIZE)
{
fprintf(stderr,"OOPS genotype file has > %d genotypes\n",MAXSIZE);
fprintf(fpout,"OOPS genotype file has > %d genotypes\n",MAXSIZE);
exit(1);
}
}
}
int main(int argc, char **argv)
{
int **snppos ;
int *snpindx ;
char **snpnamelist, **indnamelist ;
char **eglist ;
int lsnplist, lindlist, numeg ;
int i,j;
SNP *cupt, *cupt1, *cupt2, *cupt3 ;
Indiv *indx ;
double gpos1,gpos2,cpos1,cpos2,gd, cd, gd100 ;
double rthresh, zt ;
int mpflag, ret, nvalid;
int ch1, ch2 ;
int fmnum , lmnum ;
int num, n1, n2 ;
int nkill = 0 ;
int t, k ;
int nindiv = 0, e, f, lag=1 ;
double xc[9], xd[4], xc2[9] ;
double ychi, zscore, zthresh = 20.0 ;
double y1, y2 ;
int nignore, numrisks = 1 ;
char **genolist ;
int numgenolist ;
int maxmiss ;
malexhet = YES ; // convertf default is don't change the data
tersem = YES ; // no snp counts
readcommands(argc, argv) ;
setomode(&outputmode, omode) ;
packmode = YES ;
settersemode(tersem) ;
numsnps =
getsnps(snpname, &snpmarkers, 0.0, badsnpname, &nignore, numrisks) ;
for (i=0; i<numsnps; i++) {
if (xchrom == -1) break ;
cupt = snpmarkers[i] ;
if (cupt -> chrom != xchrom) cupt -> ignore = YES ;
if (cupt -> ignore) continue ;
t = nnint(cupt -> physpos) ;
if ( (t< lopos) || (t >hipos)) cupt -> ignore = YES ;
}
nignore = 0 ;
for (i=0; i<numsnps; i++) {
cupt = snpmarkers[i] ;
if (cupt -> chrom > maxchrom) cupt -> ignore = YES ;
if (cupt -> chrom < minchrom) cupt -> ignore = YES ;
if (cupt -> ignore) ++nignore ;
}
if (numsnps == nignore) fatalx("no valid snps\n") ;
/**
cupt = snpmarkers[0] ;
printf("zz2: %d %d %d %20s: %d\n", numsnps, nignore, cupt -> chrom, cupt -> ID, cupt -> ignore) ;
*/
numindivs = getindivs(indivname, &indivmarkers) ;
if (genotypelist!= NULL) {
getgenos_list(genotypelist, snpmarkers, indivmarkers,
numsnps, numindivs, nignore) ;
}
else {
setgenotypename(&genotypename, indivname) ;
getgenos(genotypename, snpmarkers, indivmarkers,
numsnps, numindivs, nignore) ;
}
if (outputall) {
outfiles(snpoutfilename, indoutfilename, genooutfilename,
snpmarkers, indivmarkers, numsnps, numindivs, packout, ogmode) ;
printf("##end of convertf run (outputall mode)\n") ;
return 0 ;
}
if (killr2) {
nkill = killhir2(snpmarkers, numsnps, numindivs, r2physlim, r2genlim, r2thresh) ;
if (nkill>0) printf("killhir2. number of snps killed: %d\n", nkill) ;
}
setstatus(indivmarkers, numindivs, "Case") ;
/******************************************************************/
/* removesubthreshold(indivmarkers, snpmarkers, numindiv, numsnps,
maxmissfracind, maxmissfracsnp); */
/******************************************************************/
if (fastdup) {
if (fastdupnum > 0) setfastdupnum(fastdupnum) ;
fastdupcheck(snpmarkers, indivmarkers, numsnps, numindivs) ;
}
if (decim>0) {
snpdecimate(snpmarkers, numsnps, decim, dmindis, dmaxdis) ;
}
outfiles(snpoutfilename, indoutfilename, genooutfilename,
snpmarkers, indivmarkers, numsnps, numindivs, packout, ogmode) ;
printf("##end of convertf run\n") ;
return 0 ;
}
