static int32_t countOverlapsEntry(GTFtree *t, GTFentry *e, uint32_t start, uint32_t end, int strand, int matchType, int strandType, int direction, int32_t max, FILTER_ENTRY_FUNC ffunc) {
int dir;
int32_t cnt = 0;
if(!e) return cnt;
switch(matchType) {
case GTF_MATCH_EXACT :
if((dir = rangeExact(start, end, e)) == 0) {
cnt = 1;
}
break;
case GTF_MATCH_WITHIN :
if((dir = rangeAny(start, end, e)) == 0) {
if(rangeWithin(start, end, e) == 0) cnt = 1;
}
break;
case GTF_MATCH_CONTAIN :
if((dir = rangeAny(start, end, e)) == 0) {
if(rangeContains(start, end, e) == 0) cnt = 1;
}
break;
case GTF_MATCH_START :
if((dir = rangeStart(start, end, e)) == 0) {
cnt = 1;
}
break;
case GTF_MATCH_END :
if((dir = rangeEnd(start, end, e)) == 0) {
cnt = 1;
}
break;
default :
if((dir = rangeAny(start, end, e)) == 0) {
cnt = 1;
}
break;
}
if(cnt) {
if(!matchingStrand(e, strand, strandType)) cnt = 0;
}
if(cnt && ffunc) {
if(!ffunc(t, e)) cnt = 0;
}
if(max && cnt >= max) return max;
if(direction) {
if(dir > 0) return cnt;
return cnt + countOverlapsEntry(t, e->right, start, end, strand, matchType, strandType, direction, max, ffunc);
} else {
if(dir < 0) return cnt;
return cnt + countOverlapsEntry(t, e->left, start, end, strand, matchType, strandType, direction, max, ffunc);
}
}
void mergeAll()
/* read through nameHash */
/* look up in 4 population hashes */
/* call logMissing to report on missing data */
/* check chrom, position, strand, observed */
{
char outputFileName[64];
FILE *outputFileHandle = NULL;
struct hashCookie cookie;
struct hashEl *nameHashElement = NULL;
struct hashEl *helCEU = NULL;
struct hashEl *helCHB = NULL;
struct hashEl *helJPT = NULL;
struct hashEl *helYRI = NULL;
boolean allMatch = TRUE;
struct hapmapSnpsCombined *hap = NULL;
safef(outputFileName, sizeof(outputFileName), "hapmapSnpsCombined.tab");
outputFileHandle = mustOpen(outputFileName, "w");
cookie = hashFirst(nameHash);
while ((nameHashElement = hashNext(&cookie)) != NULL)
{
helCEU = hashLookup(hashCEU, nameHashElement->name);
helCHB = hashLookup(hashCHB, nameHashElement->name);
helJPT = hashLookup(hashJPT, nameHashElement->name);
helYRI = hashLookup(hashYRI, nameHashElement->name);
/* should convert to instance of hapmapSnps here */
logMissing(nameHashElement->name, helCEU, helCHB, helJPT, helYRI);
allMatch = matchingChroms(helCEU, helCHB, helJPT, helYRI);
if (!allMatch)
{
fprintf(errorFileHandle, "different chroms for %s\n", nameHashElement->name);
continue;
}
allMatch = matchingPositions(helCEU, helCHB, helJPT, helYRI);
if (!allMatch)
{
fprintf(errorFileHandle, "different positions for %s\n", nameHashElement->name);
continue;
}
/* if strand is different, log and fix if possible */
allMatch = matchingStrand(helCEU, helCHB, helJPT, helYRI);
if (!allMatch)
{
fprintf(errorFileHandle, "different strands for %s\n", nameHashElement->name);
hap = fixStrandAndMerge(helCEU, helCHB, helJPT, helYRI);
if (hap)
writeOutput(hap, outputFileHandle);
continue;
}
/* different observed is not a fatal error? */
allMatch = matchingObserved(helCEU, helCHB, helJPT, helYRI);
if (!allMatch)
{
fprintf(errorFileHandle, "different observed for %s\n", nameHashElement->name);
continue;
}
/* should just have 2 alleles in all */
allMatch = matchingAlleles(helCEU, helCHB, helJPT, helYRI);
if (!allMatch)
{
fprintf(errorFileHandle, "different alleles for %s\n", nameHashElement->name);
continue;
}
hap = mergeOne(helCEU, helCHB, helJPT, helYRI);
if (hap)
writeOutput(hap, outputFileHandle);
}
carefulClose(&outputFileHandle);
}
