struct chromAnnMap *chromAnnMapNew()
/* construct a new object */
{
struct chromAnnMap *cam;
AllocVar(cam);
cam->ranges = genomeRangeTreeNew();
return cam;
}
void edwBamToWig(char *input, char *output)
/* edwBamToWig - Convert a bam file to a wig file by measuring depth of coverage, optionally adjusting hit size to average for library.. */
{
FILE *f = mustOpen(output, "w");
/* Open file and get header for it. */
samfile_t *sf = samopen(input, "rb", NULL);
if (sf == NULL)
errnoAbort("Couldn't open %s.\n", input);
bam_header_t *head = sf->header;
if (head == NULL)
errAbort("Aborting ... Bad BAM header in file: %s", input);
/* Scan through input populating genome range trees */
struct genomeRangeTree *grt = genomeRangeTreeNew();
bam1_t one = {};
for (;;)
{
/* Read next record. */
if (bam_read1(sf->x.bam, &one) < 0)
break;
if (one.core.tid >= 0 && one.core.n_cigar > 0)
{
char *chrom = head->target_name[one.core.tid];
int start = one.core.pos;
int end = start + one.core.l_qseq;
if (one.core.flag & BAM_FREVERSE)
{
start -= clPad;
}
else
{
end += clPad;
}
struct rbTree *rt = genomeRangeTreeFindOrAddRangeTree(grt,chrom);
rangeTreeAddToCoverageDepth(rt, start, end);
}
}
/* Convert genome range tree into output wig */
/* Get list of chromosomes. */
struct hashEl *hel, *helList = hashElListHash(grt->hash);
for (hel = helList; hel != NULL; hel = hel->next)
{
char *chrom = hel->name;
struct rbTree *rt = hel->val;
struct range *range, *rangeList = rangeTreeList(rt);
for (range = rangeList; range != NULL; range = range->next)
{
fprintf(f, "%s\t%d\t%d\t%d\n", chrom, range->start, range->end, ptToInt(range->val));
}
}
carefulClose(&f);
}
struct genomeRangeTree *edwMakeGrtFromBed3List(struct bed3 *bedList)
/* Make up a genomeRangeTree around bed file. */
{
struct genomeRangeTree *grt = genomeRangeTreeNew();
struct bed3 *bed;
for (bed = bedList; bed != NULL; bed = bed->next)
genomeRangeTreeAdd(grt, bed->chrom, bed->chromStart, bed->chromEnd);
return grt;
}
/* build the range tree when needed */
static void buildRangeTree(struct malnSet *malnSet) {
malnSet->compRangeMap = genomeRangeTreeNew();
struct malnBlkSetIterator *iter = malnBlkSet_getIterator(malnSet->blks);
struct malnBlk *blk;
while ((blk = malnBlkSetIterator_getNext(iter)) != NULL) {
addCompsToMap(malnSet, blk);
}
malnBlkSetIterator_destruct(iter);
}
struct genomeRangeTree* getRangeTreeOfRegdoms(struct regdom* regdoms)
{
struct genomeRangeTree *ranges = genomeRangeTreeNew();
struct regdom* currRD;
for (currRD = regdoms; currRD != NULL; currRD = currRD->next) {
genomeRangeTreeAdd(ranges, currRD->chrom, currRD->chromStart, currRD->chromEnd);
}
return ranges;
}
static void subset_with_sections(struct metaBig* mb, struct bed** p_list)
/* mainly for chopgenome */
{
struct genomeRangeTree* grt = genomeRangeTreeNew();
struct bed* sec;
struct bed* list;
struct bed* newlist = NULL;
struct bed* head;
for (sec = mb->sections; sec != NULL; sec = sec->next)
genomeRangeTreeAdd(grt, sec->chrom, sec->chromStart, sec->chromEnd);
list = *p_list;
while ((head = slPopHead(&list)) != NULL) {
if (genomeRangeTreeOverlaps(grt, head->chrom, head->chromStart, head->chromEnd) && genomeRangeTreeFindEnclosing(grt, head->chrom, head->chromStart, head->chromEnd))
slAddHead(&newlist, head);
else
bedFree(&head);
}
slReverse(&newlist);
*p_list = newlist;
}
struct genomeRangeTree *edwGrtFromBigBed(char *fileName)
/* Return genome range tree for simple (unblocked) bed */
{
struct bbiFile *bbi = bigBedFileOpen(fileName);
struct bbiChromInfo *chrom, *chromList = bbiChromList(bbi);
struct genomeRangeTree *grt = genomeRangeTreeNew();
for (chrom = chromList; chrom != NULL; chrom = chrom->next)
{
struct rbTree *tree = genomeRangeTreeFindOrAddRangeTree(grt, chrom->name);
struct lm *lm = lmInit(0);
struct bigBedInterval *iv, *ivList = NULL;
ivList = bigBedIntervalQuery(bbi, chrom->name, 0, chrom->size, 0, lm);
for (iv = ivList; iv != NULL; iv = iv->next)
rangeTreeAdd(tree, iv->start, iv->end);
lmCleanup(&lm);
}
bigBedFileClose(&bbi);
bbiChromInfoFreeList(&chromList);
return grt;
}
struct genomeRangeTree *grtFromOpenBed(struct lineFile *lf, int size, boolean doPromoter)
/* Read an open bed file into a genomeRangeTree and return it. */
{
struct genomeRangeTree *grt = genomeRangeTreeNew();
char *row[size];
while (lineFileRow(lf, row))
{
struct bed *bed = bedLoadN(row, size);
if (doPromoter)
{
if (bed->strand[0] == '+')
genomeRangeTreeAdd(grt, bed->chrom, bed->chromStart - 500, bed->chromEnd + 500);
else
genomeRangeTreeAdd(grt, bed->chrom, bed->chromEnd - 500, bed->chromEnd + 500);
}
else
genomeRangeTreeAdd(grt, bed->chrom, bed->chromStart, bed->chromEnd);
}
return grt;
}
void edwAlignFastqMakeBed(struct edwFile *ef, struct edwAssembly *assembly,
char *fastqPath, struct edwValidFile *vf, FILE *bedF,
double *retMapRatio, double *retDepth, double *retSampleCoverage)
/* Take a sample fastq and run bwa on it, and then convert that file to a bed.
* bedF and all the ret parameters can be NULL. */
{
/* Hmm, tried doing this with Mark's pipeline code, but somehow it would be flaky the
* second time it was run in same app. Resorting therefore to temp files. */
char genoFile[PATH_LEN];
safef(genoFile, sizeof(genoFile), "%s%s/bwaData/%s.fa",
edwValDataDir, assembly->ucscDb, assembly->ucscDb);
char cmd[3*PATH_LEN];
char *saiName = cloneString(rTempName(edwTempDir(), "edwSample1", ".sai"));
safef(cmd, sizeof(cmd), "bwa aln -t 3 %s %s > %s", genoFile, fastqPath, saiName);
mustSystem(cmd);
char *samName = cloneString(rTempName(edwTempDir(), "ewdSample1", ".sam"));
safef(cmd, sizeof(cmd), "bwa samse %s %s %s > %s", genoFile, saiName, fastqPath, samName);
mustSystem(cmd);
remove(saiName);
/* Scan sam file to calculate vf->mapRatio, vf->sampleCoverage and vf->depth.
* and also to produce little bed file for enrichment step. */
struct genomeRangeTree *grt = genomeRangeTreeNew();
long long hitCount=0, missCount=0, totalBasesInHits=0;
scanSam(samName, bedF, grt, &hitCount, &missCount, &totalBasesInHits);
verbose(1, "hitCount=%lld, missCount=%lld, totalBasesInHits=%lld, grt=%p\n",
hitCount, missCount, totalBasesInHits, grt);
if (retMapRatio)
*retMapRatio = (double)hitCount/(hitCount+missCount);
if (retDepth)
*retDepth = (double)totalBasesInHits/assembly->baseCount
* (double)vf->itemCount/vf->sampleCount;
long long basesHitBySample = genomeRangeTreeSumRanges(grt);
if (retSampleCoverage)
*retSampleCoverage = (double)basesHitBySample/assembly->baseCount;
genomeRangeTreeFree(&grt);
remove(samName);
}
