void calculate_mq_params(bamFile* const pfp,
bam_index_t* const fp_index,
const int32_t tid,
const int32_t start,
const int32_t stop,
uint* const mqrms,
uint* const mq0)
{
bamFile fp = *pfp;
mapqcov* mqds = ckallocz(sizeof(mapqcov));
bam_fetch(fp,
fp_index,
tid, start, stop,
&mqds,
calcrms);
*mqrms = sqrt(mqds->mqsqsum*1.0/mqds->numcov);
*mq0 = mqds->numcov0;
ckfree(mqds);
return;
}
bool YTranscriptFetcher::fetchBAMTranscripts(const char* filename, const char *refName, unsigned int start, unsigned int end, std::vector<YTranscript*> *transcripts,std::set<std::string> *transcriptNames) {
//Open the region in the bam file
fetch_data_t data;
fetch_data_t *d = &data;
d->beg = start-1-buffer;
d->end = end+buffer;
d->transcripts = transcripts;
d->requestedTranscripts = transcriptNames;
d->in = samopen(filename, "rb", 0);
if (d->in == 0) {
fprintf(stderr, "Failed to open BAM file %s\n", filename);
return 0;
}
bam_index_t *idx;
idx = bam_index_load(filename); // load BAM index
if (idx == 0) {
fprintf(stderr, "BAM indexing file is not available.\n");
return 0;
}
bam_init_header_hash(d->in->header);
d->tid = bam_get_tid(d->in->header, refName);
if(d->tid == -1) {
fprintf(stderr, "Reference id %s not found in BAM file",refName);
return 0;
}
bam_fetch(d->in->x.bam, idx, d->tid, d->beg, d->end, d, fetch_func);
bam_index_destroy(idx);
samclose(d->in);
return 1;
}
uint calculate_cov_params(const char* const bam_name,
const int32_t tid,
const int32_t start,
const int32_t stop)
{
bamFile fp = bam_open(bam_name, "r");
bam_index_t* fp_index = bam_index_load(bam_name);
bam_plbuf_t *buf;
covdata* cvdt = ckallocz(sizeof(covdata));
cvdt->tid = tid;
cvdt->begin = start;
cvdt->end = stop;
cvdt->coverage = ckallocz((cvdt->end - cvdt->begin) * sizeof(uint32_t));
buf = bam_plbuf_init(pileup_func, cvdt);
bam_fetch(fp, fp_index, tid, start, stop, buf, fetch_func);
bam_plbuf_push(0, buf);
bam_plbuf_destroy(buf);
// calculate the mean coverage in the region of the putative deletion
uint i, covsum;
for(i = 0, covsum = 0; i < (cvdt->end - cvdt->begin); i++){
covsum += cvdt->coverage[i];
}
uint avgcov = floor(covsum * 1.0/(cvdt->end - cvdt->begin));
ckfree(cvdt->coverage);
ckfree(cvdt);
bam_close(fp);
bam_index_destroy(fp_index);
return avgcov;
}
int main(int argc, char *argv[])
{
char *progname;
char *bamfilename;
int32_t tid;
samfile_t *bamin;
bam_index_t *bamidx;
bam_plbuf_t *buf;
bam1_t *bam_read;
uint32_t next_pos = 1;
progname = *argv;
argv++; argc--;
if (argc < 2) {
printf("Usage: %s bam_file tid\n", progname);
exit(1);
}
else {
bamfilename = argv[0];
tid = strtol(argv[1], NULL, 10);
}
/* try to open bam file */
bamin = samopen(bamfilename, "rb", NULL);
if (!bamin) {
fprintf(stderr, "Error opening bamfile %s\n", bamfilename);
exit(1);
}
/* try to open index */
bamidx = bam_index_load(bamfilename);
if (!bamidx) {
fprintf(stderr, "Error opening index for %s\n", bamfilename);
exit(1);
}
bam_read = bam_init1();
buf = bam_plbuf_init(&pileup_func, &next_pos);
/* disable maximum pileup depth */
bam_plp_set_maxcnt(buf->iter, INT_MAX);
bam_fetch(bamin->x.bam, bamidx,
tid, 0, INT_MAX,
buf, &fetch_func);
bam_plbuf_push(0, buf); /* finish pileup */
bam_plbuf_destroy(buf);
bam_destroy1(bam_read);
bam_index_destroy(bamidx);
samclose(bamin);
return 0;
}
static int load_discordant_reads(MEI_data& mei_data, std::vector<bam_info>& bam_sources, const std::string& chr_name,
const SearchWindow& window, UserDefinedSettings* userSettings) {
// Loop over associated bam files.
for (size_t i = 0; i < bam_sources.size(); i++) {
// Locate file.
bam_info source = bam_sources.at(i);
LOG_DEBUG(*logStream << time_log() << "Loading discordant reads from " << source.BamFile << std::endl);
// Setup link to bamfile, its index and header.
bamFile fp = bam_open(source.BamFile.c_str(), "r");
bam_index_t *idx = bam_index_load(source.BamFile.c_str());
if (idx == NULL) {
LOG_WARN(*logStream << time_log() << "Failed to load index for " << source.BamFile.c_str() << std::endl);
LOG_WARN(*logStream << "Skipping window: " << chr_name << ", " << window.getStart() << "--" <<
window.getEnd() << " for BAM-file: " << source.BamFile.c_str() << std::endl);
continue;
}
bam_header_t *header = bam_header_read(fp);
bam_init_header_hash(header);
int tid = bam_get_tid(header, chr_name.c_str());
if (tid < 0) {
LOG_WARN(*logStream << time_log() << "Could not find sequence in alignment file: '" << chr_name <<
"'" << std::endl);
LOG_WARN(*logStream << "Skipping window: " << chr_name << ", " << window.getStart() << "--" <<
window.getEnd() << " for BAM-file: " << source.BamFile.c_str() << std::endl);
continue;
}
mei_data.sample_names = get_sample_dictionary(header);
// Save insert size of current bamfile in data object provided for callback function.
// Note: the insert size should ideally be separate from the MEI_data object, tried to do
// this using a std::pair object, which did not work. Suggestions are welcome here.
mei_data.current_insert_size = source.InsertSize;
mei_data.current_chr_name = chr_name;
// Set up environment variable for callback function.
std::pair<MEI_data*, UserDefinedSettings*> env = std::make_pair(&mei_data, userSettings);
// Load discordant reads into mei_data.
bam_fetch(fp, idx, tid, window.getStart(), window.getEnd(), &env, fetch_disc_read_callback);
bam_index_destroy(idx);
}
return 0;
}
void bamFetchAlreadyOpen(samfile_t *samfile, bam_index_t *idx, char *bamFileName,
char *position, bam_fetch_f callbackFunc, void *callbackData)
/* With the open bam file, return items the same way with the callbacks as with bamFetch() */
/* except in this case use an already-open bam file and index (use bam_index_load and free() for */
/* the index). It seems a little strange to pass the filename in with the open bam, but */
/* it's just used to report errors. */
{
int chromId, start, end;
int ret = bam_parse_region(samfile->header, position, &chromId, &start, &end);
if (ret != 0 && startsWith("chr", position))
ret = bam_parse_region(samfile->header, position+strlen("chr"), &chromId, &start, &end);
if (ret != 0)
// If the bam file does not cover the current chromosome, OK
return;
ret = bam_fetch(samfile->x.bam, idx, chromId, start, end, callbackData, callbackFunc);
if (ret != 0)
warn("bam_fetch(%s, %s (chromId=%d) failed (%d)", bamFileName, position, chromId, ret);
}
int main(int argc, char *argv[])
{
tmpstruct_t tmp;
if (argc == 1) {
fprintf(stderr, "Usage: calDepth <in.bam> [region]\n");
return 1;
}
tmp.beg = 0; tmp.end = 0x7fffffff;
tmp.in = samopen(argv[1], "rb", 0);
if (tmp.in == 0) {
fprintf(stderr, "Fail to open BAM file %s\n", argv[1]);
return 1;
}
if (argc == 2) { // if a region is not specified
sampileup(tmp.in, -1, pileup_func, &tmp);
} else {
int ref;
bam_index_t *idx;
bam_plbuf_t *buf;
idx = bam_index_load(argv[1]); // load BAM index
if (idx == 0) {
fprintf(stderr, "BAM indexing file is not available.\n");
return 1;
}
bam_parse_region(tmp.in->header, argv[2], &ref,
&tmp.beg, &tmp.end); // parse the region
if (ref < 0) {
fprintf(stderr, "Invalid region %s\n", argv[2]);
return 1;
}
buf = bam_plbuf_init(pileup_func, &tmp); // initialize pileup
bam_fetch(tmp.in->x.bam, idx, ref, tmp.beg, tmp.end, buf, fetch_func);
bam_plbuf_push(0, buf); // finalize pileup
bam_index_destroy(idx);
bam_plbuf_destroy(buf);
}
samclose(tmp.in);
return 0;
}
