bufReader initBufReader2(const char*fname,int doCheck,char *fai_fname){
bufReader ret;
ret.fn = strdup(fname);
int newlen=strlen(fname);//<-just to avoid valgrind -O3 uninitialized warning
ret.fp = openBAM(ret.fn,doCheck);
if (fai_fname && hts_set_fai_filename(ret.fp, fai_fname) != 0) {
fprintf(stderr, "[%s] failed to process %s\n",
__func__, fai_fname);
exit(EXIT_FAILURE);
}
ret.isEOF =0;
ret.itr=NULL;
ret.idx=NULL;
ret.hdr = sam_hdr_read(ret.fp);
if(strlen(ret.hdr->text)==0){
fprintf(stderr,"\t-> No header information could be found for BAM/CRAM file: \'%s\' will exit\n",fname);
exit(1);
}
checkIfSorted(ret.hdr->text);
if(ret.hdr==NULL) {
fprintf(stderr, "[main_samview] fail to read the header from \"%s\".\n", ret.fn);
exit(0);
}
return ret;
}
int bam_access_openhts(char *hts_file, char *ref_file){
assert(hts_file != NULL);
//Assign memory for the file name etc holding struct
fholder = malloc(sizeof(file_holder));
check_mem(fholder);
//Beginning and end of tmp struct for bam access
fholder->beg = 0; fholder->end = 0x7fffffff; // The max 32 bit integer.
//Open a file for read from compressed bam.
fholder->in = hts_open(hts_file, "r");
check(fholder->in != 0,"HTS file %s failed to open.",hts_file);
fholder->idx = sam_index_load(fholder->in,hts_file);
check(fholder->idx != 0,"HTS index file %s failed to open.",hts_file);
if(ref_file){
hts_set_fai_filename(fholder->in, ref_file);
}else{
if(fholder->in->format.format == cram) log_warn("No reference file provided for a cram input file, if the reference described in the cram header can't be located this script may fail.");
}
//Check for generic header read method.
fholder->head = sam_hdr_read(fholder->in);
return 0;
error:
if(fholder->in) hts_close(fholder->in);
if(fholder) free(fholder);
return -1;
}
samfile_t *samopen(const char *fn, const char *mode, const void *aux)
{
// hts_open() is really sam_open(), except for #define games
samFile *hts_fp = hts_open(fn, mode);
if (hts_fp == NULL) return NULL;
samfile_t *fp = malloc(sizeof (samfile_t));
if (!fp) {
sam_close(hts_fp);
return NULL;
}
fp->file = hts_fp;
fp->x.bam = hts_fp->fp.bgzf;
if (strchr(mode, 'r')) {
if (aux) {
if (hts_set_fai_filename(fp->file, aux) != 0) {
sam_close(hts_fp);
free(fp);
return NULL;
}
}
fp->header = sam_hdr_read(fp->file); // samclose() will free this
if (fp->header == NULL) {
sam_close(hts_fp);
free(fp);
return NULL;
}
fp->is_write = 0;
if (fp->header->n_targets == 0 && bam_verbose >= 1)
fprintf(samtools_stderr, "[samopen] no @SQ lines in the header.\n");
}
else {
enum htsExactFormat fmt = hts_get_format(fp->file)->format;
fp->header = (bam_hdr_t *)aux; // For writing, we won't free it
fp->is_write = 1;
if (!(fmt == text_format || fmt == sam) || strchr(mode, 'h')) {
if (sam_hdr_write(fp->file, fp->header) < 0) {
if (bam_verbose >= 1)
fprintf(samtools_stderr, "[samopen] Couldn't write header\n");
sam_close(hts_fp);
free(fp);
return NULL;
}
}
}
return fp;
}
static int mpileup(mplp_conf_t *conf)
{
if (conf->nfiles == 0) {
fprintf(stderr,"[%s] no input file/data given\n", __func__);
exit(EXIT_FAILURE);
}
mplp_ref_t mp_ref = MPLP_REF_INIT;
conf->gplp = (mplp_pileup_t *) calloc(1,sizeof(mplp_pileup_t));
conf->mplp_data = (mplp_aux_t**) calloc(conf->nfiles, sizeof(mplp_aux_t*));
conf->plp = (const bam_pileup1_t**) calloc(conf->nfiles, sizeof(bam_pileup1_t*));
conf->n_plp = (int*) calloc(conf->nfiles, sizeof(int));
// Allow to run mpileup on multiple regions in one go. This comes at cost: the bai index
// must be kept in the memory for the whole time which can be a problem with many bams.
// Therefore if none or only one region is requested, we initialize the bam iterator as
// before and free the index. Only when multiple regions are queried, we keep the index.
int nregs = 0;
if ( conf->reg_fname )
{
if ( conf->reg_is_file )
{
conf->reg = regidx_init(conf->reg_fname,NULL,NULL,0,NULL);
if ( !conf->reg ) {
fprintf(stderr,"Could not parse the regions: %s\n", conf->reg_fname);
exit(EXIT_FAILURE);
}
}
else
{
conf->reg = regidx_init(NULL,regidx_parse_reg,NULL,sizeof(char*),NULL);
if ( regidx_insert_list(conf->reg,conf->reg_fname,',') !=0 ) {
fprintf(stderr,"Could not parse the regions: %s\n", conf->reg_fname);
exit(EXIT_FAILURE);
}
}
nregs = regidx_nregs(conf->reg);
conf->reg_itr = regitr_init(conf->reg);
regitr_loop(conf->reg_itr); // region iterator now positioned at the first region
}
// read the header of each file in the list and initialize data
// beware: mpileup has always assumed that tid's are consistent in the headers, add sanity check at least!
bam_hdr_t *hdr = NULL; // header of first file in input list
int i;
for (i = 0; i < conf->nfiles; ++i) {
bam_hdr_t *h_tmp;
conf->mplp_data[i] = (mplp_aux_t*) calloc(1, sizeof(mplp_aux_t));
conf->mplp_data[i]->fp = sam_open(conf->files[i], "rb");
if ( !conf->mplp_data[i]->fp )
{
fprintf(stderr, "[%s] failed to open %s: %s\n", __func__, conf->files[i], strerror(errno));
exit(EXIT_FAILURE);
}
if (hts_set_opt(conf->mplp_data[i]->fp, CRAM_OPT_DECODE_MD, 0)) {
fprintf(stderr, "Failed to set CRAM_OPT_DECODE_MD value\n");
exit(EXIT_FAILURE);
}
if (conf->fai_fname && hts_set_fai_filename(conf->mplp_data[i]->fp, conf->fai_fname) != 0) {
fprintf(stderr, "[%s] failed to process %s: %s\n",
__func__, conf->fai_fname, strerror(errno));
exit(EXIT_FAILURE);
}
conf->mplp_data[i]->conf = conf;
conf->mplp_data[i]->ref = &mp_ref;
h_tmp = sam_hdr_read(conf->mplp_data[i]->fp);
if ( !h_tmp ) {
fprintf(stderr,"[%s] fail to read the header of %s\n", __func__, conf->files[i]);
exit(EXIT_FAILURE);
}
conf->mplp_data[i]->h = i ? hdr : h_tmp; // for j==0, "h" has not been set yet
conf->mplp_data[i]->bam_id = bam_smpl_add_bam(conf->bsmpl,h_tmp->text,conf->files[i]);
if ( conf->mplp_data[i]->bam_id<0 )
{
// no usable readgroups in this bam, it can be skipped
sam_close(conf->mplp_data[i]->fp);
free(conf->mplp_data[i]);
bam_hdr_destroy(h_tmp);
free(conf->files[i]);
if ( i+1<conf->nfiles ) memmove(&conf->files[i],&conf->files[i+1],sizeof(*conf->files)*(conf->nfiles-i-1));
conf->nfiles--;
i--;
continue;
}
if (conf->reg) {
hts_idx_t *idx = sam_index_load(conf->mplp_data[i]->fp, conf->files[i]);
if (idx == NULL) {
fprintf(stderr, "[%s] fail to load index for %s\n", __func__, conf->files[i]);
exit(EXIT_FAILURE);
}
conf->buf.l = 0;
ksprintf(&conf->buf,"%s:%u-%u",conf->reg_itr->seq,conf->reg_itr->beg+1,conf->reg_itr->end+1);
conf->mplp_data[i]->iter = sam_itr_querys(idx, conf->mplp_data[i]->h, conf->buf.s);
if ( !conf->mplp_data[i]->iter )
{
conf->mplp_data[i]->iter = sam_itr_querys(idx, conf->mplp_data[i]->h, conf->reg_itr->seq);
if ( conf->mplp_data[i]->iter ) {
fprintf(stderr,"[E::%s] fail to parse region '%s'\n", __func__, conf->buf.s);
exit(EXIT_FAILURE);
}
fprintf(stderr,"[E::%s] the sequence \"%s\" not found: %s\n",__func__,conf->reg_itr->seq,conf->files[i]);
exit(EXIT_FAILURE);
}
if ( nregs==1 ) // no need to keep the index in memory
hts_idx_destroy(idx);
else
conf->mplp_data[i]->idx = idx;
}
if ( !hdr ) hdr = h_tmp; /* save the header of first file in list */
else {
// FIXME: check consistency between h and h_tmp
bam_hdr_destroy(h_tmp);
// we store only the first file's header; it's (alleged to be)
// compatible with the i-th file's target_name lookup needs
conf->mplp_data[i]->h = hdr;
}
}
// allocate data storage proportionate to number of samples being studied sm->n
bam_smpl_get_samples(conf->bsmpl, &conf->gplp->n);
conf->gplp->n_plp = (int*) calloc(conf->gplp->n, sizeof(int));
conf->gplp->m_plp = (int*) calloc(conf->gplp->n, sizeof(int));
conf->gplp->plp = (bam_pileup1_t**) calloc(conf->gplp->n, sizeof(bam_pileup1_t*));
fprintf(stderr, "[%s] %d samples in %d input files\n", __func__, conf->gplp->n, conf->nfiles);
// write the VCF header
conf->bcf_fp = hts_open(conf->output_fname?conf->output_fname:"-", hts_bcf_wmode(conf->output_type));
if (conf->bcf_fp == NULL) {
fprintf(stderr, "[%s] failed to write to %s: %s\n", __func__, conf->output_fname? conf->output_fname : "standard output", strerror(errno));
exit(EXIT_FAILURE);
}
if ( conf->n_threads ) hts_set_threads(conf->bcf_fp, conf->n_threads);
// BCF header creation
conf->bcf_hdr = bcf_hdr_init("w");
conf->buf.l = 0;
if (conf->record_cmd_line)
{
ksprintf(&conf->buf, "##bcftoolsVersion=%s+htslib-%s\n",bcftools_version(),hts_version());
bcf_hdr_append(conf->bcf_hdr, conf->buf.s);
conf->buf.l = 0;
ksprintf(&conf->buf, "##bcftoolsCommand=mpileup");
for (i=1; i<conf->argc; i++) ksprintf(&conf->buf, " %s", conf->argv[i]);
kputc('\n', &conf->buf);
bcf_hdr_append(conf->bcf_hdr, conf->buf.s);
}
if (conf->fai_fname)
{
conf->buf.l = 0;
ksprintf(&conf->buf, "##reference=file://%s\n", conf->fai_fname);
bcf_hdr_append(conf->bcf_hdr, conf->buf.s);
}
// Translate BAM @SQ tags to BCF ##contig tags
// todo: use/write new BAM header manipulation routines, fill also UR, M5
for (i=0; i<hdr->n_targets; i++)
{
conf->buf.l = 0;
ksprintf(&conf->buf, "##contig=<ID=%s,length=%d>", hdr->target_name[i], hdr->target_len[i]);
bcf_hdr_append(conf->bcf_hdr, conf->buf.s);
}
conf->buf.l = 0;
bcf_hdr_append(conf->bcf_hdr,"##ALT=<ID=*,Description=\"Represents allele(s) other than observed.\">");
bcf_hdr_append(conf->bcf_hdr,"##INFO=<ID=INDEL,Number=0,Type=Flag,Description=\"Indicates that the variant is an INDEL.\">");
bcf_hdr_append(conf->bcf_hdr,"##INFO=<ID=IDV,Number=1,Type=Integer,Description=\"Maximum number of reads supporting an indel\">");
bcf_hdr_append(conf->bcf_hdr,"##INFO=<ID=IMF,Number=1,Type=Float,Description=\"Maximum fraction of reads supporting an indel\">");
bcf_hdr_append(conf->bcf_hdr,"##INFO=<ID=DP,Number=1,Type=Integer,Description=\"Raw read depth\">");
bcf_hdr_append(conf->bcf_hdr,"##INFO=<ID=VDB,Number=1,Type=Float,Description=\"Variant Distance Bias for filtering splice-site artefacts in RNA-seq data (bigger is better)\",Version=\"3\">");
bcf_hdr_append(conf->bcf_hdr,"##INFO=<ID=RPB,Number=1,Type=Float,Description=\"Mann-Whitney U test of Read Position Bias (bigger is better)\">");
bcf_hdr_append(conf->bcf_hdr,"##INFO=<ID=MQB,Number=1,Type=Float,Description=\"Mann-Whitney U test of Mapping Quality Bias (bigger is better)\">");
bcf_hdr_append(conf->bcf_hdr,"##INFO=<ID=BQB,Number=1,Type=Float,Description=\"Mann-Whitney U test of Base Quality Bias (bigger is better)\">");
bcf_hdr_append(conf->bcf_hdr,"##INFO=<ID=MQSB,Number=1,Type=Float,Description=\"Mann-Whitney U test of Mapping Quality vs Strand Bias (bigger is better)\">");
#if CDF_MWU_TESTS
bcf_hdr_append(conf->bcf_hdr,"##INFO=<ID=RPB2,Number=1,Type=Float,Description=\"Mann-Whitney U test of Read Position Bias [CDF] (bigger is better)\">");
bcf_hdr_append(conf->bcf_hdr,"##INFO=<ID=MQB2,Number=1,Type=Float,Description=\"Mann-Whitney U test of Mapping Quality Bias [CDF] (bigger is better)\">");
bcf_hdr_append(conf->bcf_hdr,"##INFO=<ID=BQB2,Number=1,Type=Float,Description=\"Mann-Whitney U test of Base Quality Bias [CDF] (bigger is better)\">");
bcf_hdr_append(conf->bcf_hdr,"##INFO=<ID=MQSB2,Number=1,Type=Float,Description=\"Mann-Whitney U test of Mapping Quality vs Strand Bias [CDF] (bigger is better)\">");
#endif
bcf_hdr_append(conf->bcf_hdr,"##INFO=<ID=SGB,Number=1,Type=Float,Description=\"Segregation based metric.\">");
bcf_hdr_append(conf->bcf_hdr,"##INFO=<ID=MQ0F,Number=1,Type=Float,Description=\"Fraction of MQ0 reads (smaller is better)\">");
bcf_hdr_append(conf->bcf_hdr,"##INFO=<ID=I16,Number=16,Type=Float,Description=\"Auxiliary tag used for calling, see description of bcf_callret1_t in bam2bcf.h\">");
bcf_hdr_append(conf->bcf_hdr,"##INFO=<ID=QS,Number=R,Type=Float,Description=\"Auxiliary tag used for calling\">");
bcf_hdr_append(conf->bcf_hdr,"##FORMAT=<ID=PL,Number=G,Type=Integer,Description=\"List of Phred-scaled genotype likelihoods\">");
if ( conf->fmt_flag&B2B_FMT_DP )
bcf_hdr_append(conf->bcf_hdr,"##FORMAT=<ID=DP,Number=1,Type=Integer,Description=\"Number of high-quality bases\">");
if ( conf->fmt_flag&B2B_FMT_DV )
bcf_hdr_append(conf->bcf_hdr,"##FORMAT=<ID=DV,Number=1,Type=Integer,Description=\"Number of high-quality non-reference bases\">");
if ( conf->fmt_flag&B2B_FMT_DPR )
bcf_hdr_append(conf->bcf_hdr,"##FORMAT=<ID=DPR,Number=R,Type=Integer,Description=\"Number of high-quality bases observed for each allele\">");
if ( conf->fmt_flag&B2B_INFO_DPR )
bcf_hdr_append(conf->bcf_hdr,"##INFO=<ID=DPR,Number=R,Type=Integer,Description=\"Number of high-quality bases observed for each allele\">");
if ( conf->fmt_flag&B2B_FMT_DP4 )
bcf_hdr_append(conf->bcf_hdr,"##FORMAT=<ID=DP4,Number=4,Type=Integer,Description=\"Number of high-quality ref-fwd, ref-reverse, alt-fwd and alt-reverse bases\">");
if ( conf->fmt_flag&B2B_FMT_SP )
bcf_hdr_append(conf->bcf_hdr,"##FORMAT=<ID=SP,Number=1,Type=Integer,Description=\"Phred-scaled strand bias P-value\">");
if ( conf->fmt_flag&B2B_FMT_AD )
bcf_hdr_append(conf->bcf_hdr,"##FORMAT=<ID=AD,Number=R,Type=Integer,Description=\"Allelic depths\">");
if ( conf->fmt_flag&B2B_FMT_ADF )
bcf_hdr_append(conf->bcf_hdr,"##FORMAT=<ID=ADF,Number=R,Type=Integer,Description=\"Allelic depths on the forward strand\">");
if ( conf->fmt_flag&B2B_FMT_ADR )
bcf_hdr_append(conf->bcf_hdr,"##FORMAT=<ID=ADR,Number=R,Type=Integer,Description=\"Allelic depths on the reverse strand\">");
if ( conf->fmt_flag&B2B_INFO_AD )
bcf_hdr_append(conf->bcf_hdr,"##INFO=<ID=AD,Number=R,Type=Integer,Description=\"Total allelic depths\">");
if ( conf->fmt_flag&B2B_INFO_ADF )
bcf_hdr_append(conf->bcf_hdr,"##INFO=<ID=ADF,Number=R,Type=Integer,Description=\"Total allelic depths on the forward strand\">");
if ( conf->fmt_flag&B2B_INFO_ADR )
bcf_hdr_append(conf->bcf_hdr,"##INFO=<ID=ADR,Number=R,Type=Integer,Description=\"Total allelic depths on the reverse strand\">");
if ( conf->gvcf )
gvcf_update_header(conf->gvcf, conf->bcf_hdr);
int nsmpl;
const char **smpl = bam_smpl_get_samples(conf->bsmpl, &nsmpl);
for (i=0; i<nsmpl; i++)
bcf_hdr_add_sample(conf->bcf_hdr, smpl[i]);
bcf_hdr_write(conf->bcf_fp, conf->bcf_hdr);
conf->bca = bcf_call_init(-1., conf->min_baseQ);
conf->bcr = (bcf_callret1_t*) calloc(nsmpl, sizeof(bcf_callret1_t));
conf->bca->openQ = conf->openQ, conf->bca->extQ = conf->extQ, conf->bca->tandemQ = conf->tandemQ;
conf->bca->min_frac = conf->min_frac;
conf->bca->min_support = conf->min_support;
conf->bca->per_sample_flt = conf->flag & MPLP_PER_SAMPLE;
conf->bc.bcf_hdr = conf->bcf_hdr;
conf->bc.n = nsmpl;
conf->bc.PL = (int32_t*) malloc(15 * nsmpl * sizeof(*conf->bc.PL));
if (conf->fmt_flag)
{
assert( sizeof(float)==sizeof(int32_t) );
conf->bc.DP4 = (int32_t*) malloc(nsmpl * sizeof(int32_t) * 4);
conf->bc.fmt_arr = (uint8_t*) malloc(nsmpl * sizeof(float)); // all fmt_flag fields, float and int32
if ( conf->fmt_flag&(B2B_INFO_DPR|B2B_FMT_DPR|B2B_INFO_AD|B2B_INFO_ADF|B2B_INFO_ADR|B2B_FMT_AD|B2B_FMT_ADF|B2B_FMT_ADR) )
{
// first B2B_MAX_ALLELES fields for total numbers, the rest per-sample
conf->bc.ADR = (int32_t*) malloc((nsmpl+1)*B2B_MAX_ALLELES*sizeof(int32_t));
conf->bc.ADF = (int32_t*) malloc((nsmpl+1)*B2B_MAX_ALLELES*sizeof(int32_t));
for (i=0; i<nsmpl; i++)
{
conf->bcr[i].ADR = conf->bc.ADR + (i+1)*B2B_MAX_ALLELES;
conf->bcr[i].ADF = conf->bc.ADF + (i+1)*B2B_MAX_ALLELES;
}
}
}
// init mpileup
conf->iter = bam_mplp_init(conf->nfiles, mplp_func, (void**)conf->mplp_data);
if ( conf->flag & MPLP_SMART_OVERLAPS ) bam_mplp_init_overlaps(conf->iter);
if ( (double)conf->max_depth * conf->nfiles > 1<<20)
fprintf(stderr, "Warning: Potential memory hog, up to %.0fM reads in the pileup!\n", (double)conf->max_depth*conf->nfiles);
if ( (double)conf->max_depth * conf->nfiles / nsmpl < 250 )
fprintf(stderr, "Note: The maximum per-sample depth with -d %d is %.1fx\n", conf->max_depth,(double)conf->max_depth * conf->nfiles / nsmpl);
bam_mplp_set_maxcnt(conf->iter, conf->max_depth);
conf->max_indel_depth = conf->max_indel_depth * nsmpl;
conf->bcf_rec = bcf_init1();
bam_mplp_constructor(conf->iter, pileup_constructor);
// Run mpileup for multiple regions
if ( nregs )
{
int ireg = 0;
do
{
// first region is already positioned
if ( ireg++ > 0 )
{
conf->buf.l = 0;
ksprintf(&conf->buf,"%s:%u-%u",conf->reg_itr->seq,conf->reg_itr->beg,conf->reg_itr->end);
for (i=0; i<conf->nfiles; i++)
{
hts_itr_destroy(conf->mplp_data[i]->iter);
conf->mplp_data[i]->iter = sam_itr_querys(conf->mplp_data[i]->idx, conf->mplp_data[i]->h, conf->buf.s);
if ( !conf->mplp_data[i]->iter )
{
conf->mplp_data[i]->iter = sam_itr_querys(conf->mplp_data[i]->idx, conf->mplp_data[i]->h, conf->reg_itr->seq);
if ( conf->mplp_data[i]->iter ) {
fprintf(stderr,"[E::%s] fail to parse region '%s'\n", __func__, conf->buf.s);
exit(EXIT_FAILURE);
}
fprintf(stderr,"[E::%s] the sequence \"%s\" not found: %s\n",__func__,conf->reg_itr->seq,conf->files[i]);
exit(EXIT_FAILURE);
}
bam_mplp_reset(conf->iter);
}
}
mpileup_reg(conf,conf->reg_itr->beg,conf->reg_itr->end);
}
while ( regitr_loop(conf->reg_itr) );
}
else
mpileup_reg(conf,0,0);
flush_bcf_records(conf, conf->bcf_fp, conf->bcf_hdr, NULL);
// clean up
free(conf->bc.tmp.s);
bcf_destroy1(conf->bcf_rec);
if (conf->bcf_fp)
{
hts_close(conf->bcf_fp);
bcf_hdr_destroy(conf->bcf_hdr);
bcf_call_destroy(conf->bca);
free(conf->bc.PL);
free(conf->bc.DP4);
free(conf->bc.ADR);
free(conf->bc.ADF);
free(conf->bc.fmt_arr);
free(conf->bcr);
}
if ( conf->gvcf ) gvcf_destroy(conf->gvcf);
free(conf->buf.s);
for (i = 0; i < conf->gplp->n; ++i) free(conf->gplp->plp[i]);
free(conf->gplp->plp); free(conf->gplp->n_plp); free(conf->gplp->m_plp); free(conf->gplp);
bam_mplp_destroy(conf->iter);
bam_hdr_destroy(hdr);
for (i = 0; i < conf->nfiles; ++i) {
if ( nregs>1 ) hts_idx_destroy(conf->mplp_data[i]->idx);
sam_close(conf->mplp_data[i]->fp);
if ( conf->mplp_data[i]->iter) hts_itr_destroy(conf->mplp_data[i]->iter);
free(conf->mplp_data[i]);
}
if ( conf->reg_itr ) regitr_destroy(conf->reg_itr);
free(conf->mplp_data); free(conf->plp); free(conf->n_plp);
free(mp_ref.ref[0]);
free(mp_ref.ref[1]);
return 0;
}
int main_samview(int argc, char *argv[])
{
int index;
for(index = 0; index < argc; index++) {
printf("The %d is %s\n",index,argv[index]);
}
getchar();return 0;
int c, is_header = 0, is_header_only = 0, ret = 0, compress_level = -1, is_count = 0;
int is_long_help = 0, n_threads = 0;
int64_t count = 0;
samFile *in = 0, *out = 0, *un_out=0;
bam_hdr_t *header = NULL;
char out_mode[5], out_un_mode[5], *out_format = "";
char *fn_in = 0, *fn_out = 0, *fn_list = 0, *q, *fn_un_out = 0;
sam_global_args ga = SAM_GLOBAL_ARGS_INIT;
samview_settings_t settings = {
.rghash = NULL,
.min_mapQ = 0,
.flag_on = 0,
.flag_off = 0,
.min_qlen = 0,
.remove_B = 0,
.subsam_seed = 0,
.subsam_frac = -1.,
.library = NULL,
.bed = NULL,
};
static const struct option lopts[] = {
SAM_OPT_GLOBAL_OPTIONS('-', 0, 'O', 0, 'T'),
{ "threads", required_argument, NULL, '@' },
{ NULL, 0, NULL, 0 }
};
/* parse command-line options */
strcpy(out_mode, "w");
strcpy(out_un_mode, "w");
while ((c = getopt_long(argc, argv,
"SbBcCt:h1Ho:O:q:f:F:ul:r:?T:R:L:s:@:m:x:U:",
lopts, NULL)) >= 0) {
switch (c) {
case 's':
if ((settings.subsam_seed = strtol(optarg, &q, 10)) != 0) {
srand(settings.subsam_seed);
settings.subsam_seed = rand();
}
settings.subsam_frac = strtod(q, &q);
break;
case 'm': settings.min_qlen = atoi(optarg); break;
case 'c': is_count = 1; break;
case 'S': break;
case 'b': out_format = "b"; break;
case 'C': out_format = "c"; break;
case 't': fn_list = strdup(optarg); break;
case 'h': is_header = 1; break;
case 'H': is_header_only = 1; break;
case 'o': fn_out = strdup(optarg); break;
case 'U': fn_un_out = strdup(optarg); break;
case 'f': settings.flag_on |= strtol(optarg, 0, 0); break;
case 'F': settings.flag_off |= strtol(optarg, 0, 0); break;
case 'q': settings.min_mapQ = atoi(optarg); break;
case 'u': compress_level = 0; break;
case '1': compress_level = 1; break;
case 'l': settings.library = strdup(optarg); break;
case 'L':
if ((settings.bed = bed_read(optarg)) == NULL) {
print_error_errno("view", "Could not read file \"%s\"", optarg);
ret = 1;
goto view_end;
}
break;
case 'r':
if (add_read_group_single("view", &settings, optarg) != 0) {
ret = 1;
goto view_end;
}
break;
case 'R':
if (add_read_groups_file("view", &settings, optarg) != 0) {
ret = 1;
goto view_end;
}
break;
/* REMOVED as htslib doesn't support this
//case 'x': out_format = "x"; break;
//case 'X': out_format = "X"; break;
*/
case '?': is_long_help = 1; break;
case 'B': settings.remove_B = 1; break;
case '@': n_threads = strtol(optarg, 0, 0); break;
case 'x':
{
if (strlen(optarg) != 2) {
fprintf(stderr, "main_samview: Error parsing -x auxiliary tags should be exactly two characters long.\n");
return usage(stderr, EXIT_FAILURE, is_long_help);
}
settings.remove_aux = (char**)realloc(settings.remove_aux, sizeof(char*) * (++settings.remove_aux_len));
settings.remove_aux[settings.remove_aux_len-1] = optarg;
}
break;
default:
if (parse_sam_global_opt(c, optarg, lopts, &ga) != 0)
return usage(stderr, EXIT_FAILURE, is_long_help);
break;
}
}
if (compress_level >= 0 && !*out_format) out_format = "b";
if (is_header_only) is_header = 1;
// File format auto-detection first
if (fn_out) sam_open_mode(out_mode+1, fn_out, NULL);
if (fn_un_out) sam_open_mode(out_un_mode+1, fn_un_out, NULL);
// Overridden by manual -b, -C
if (*out_format)
out_mode[1] = out_un_mode[1] = *out_format;
out_mode[2] = out_un_mode[2] = '\0';
// out_(un_)mode now 1 or 2 bytes long, followed by nul.
if (compress_level >= 0) {
char tmp[2];
tmp[0] = compress_level + '0'; tmp[1] = '\0';
strcat(out_mode, tmp);
strcat(out_un_mode, tmp);
}
if (argc == optind && isatty(STDIN_FILENO)) return usage(stdout, EXIT_SUCCESS, is_long_help); // potential memory leak...
fn_in = (optind < argc)? argv[optind] : "-";
// generate the fn_list if necessary
if (fn_list == 0 && ga.reference) fn_list = samfaipath(ga.reference);
// open file handlers
if ((in = sam_open_format(fn_in, "r", &ga.in)) == 0) {
print_error_errno("view", "failed to open \"%s\" for reading", fn_in);
ret = 1;
goto view_end;
}
if (fn_list) {
if (hts_set_fai_filename(in, fn_list) != 0) {
fprintf(stderr, "[main_samview] failed to use reference \"%s\".\n", fn_list);
ret = 1;
goto view_end;
}
}
if ((header = sam_hdr_read(in)) == 0) {
fprintf(stderr, "[main_samview] fail to read the header from \"%s\".\n", fn_in);
ret = 1;
goto view_end;
}
if (settings.rghash) { // FIXME: I do not know what "bam_header_t::n_text" is for...
char *tmp;
int l;
tmp = drop_rg(header->text, settings.rghash, &l);
free(header->text);
header->text = tmp;
header->l_text = l;
}
if (!is_count) {
if ((out = sam_open_format(fn_out? fn_out : "-", out_mode, &ga.out)) == 0) {
print_error_errno("view", "failed to open \"%s\" for writing", fn_out? fn_out : "standard output");
ret = 1;
goto view_end;
}
if (fn_list) {
if (hts_set_fai_filename(out, fn_list) != 0) {
fprintf(stderr, "[main_samview] failed to use reference \"%s\".\n", fn_list);
ret = 1;
goto view_end;
}
}
if (*out_format || is_header ||
out_mode[1] == 'b' || out_mode[1] == 'c' ||
(ga.out.format != sam && ga.out.format != unknown_format)) {
if (sam_hdr_write(out, header) != 0) {
fprintf(stderr, "[main_samview] failed to write the SAM header\n");
ret = 1;
goto view_end;
}
}
if (fn_un_out) {
if ((un_out = sam_open_format(fn_un_out, out_un_mode, &ga.out)) == 0) {
print_error_errno("view", "failed to open \"%s\" for writing", fn_un_out);
ret = 1;
goto view_end;
}
if (fn_list) {
if (hts_set_fai_filename(un_out, fn_list) != 0) {
fprintf(stderr, "[main_samview] failed to use reference \"%s\".\n", fn_list);
ret = 1;
goto view_end;
}
}
if (*out_format || is_header ||
out_un_mode[1] == 'b' || out_un_mode[1] == 'c' ||
(ga.out.format != sam && ga.out.format != unknown_format)) {
if (sam_hdr_write(un_out, header) != 0) {
fprintf(stderr, "[main_samview] failed to write the SAM header\n");
ret = 1;
goto view_end;
}
}
}
}
if (n_threads > 1) { if (out) hts_set_threads(out, n_threads); }
if (is_header_only) goto view_end; // no need to print alignments
if (optind + 1 >= argc) { // convert/print the entire file
bam1_t *b = bam_init1();
int r;
while ((r = sam_read1(in, header, b)) >= 0) { // read one alignment from `in'
if (!process_aln(header, b, &settings)) {
if (!is_count) { if (check_sam_write1(out, header, b, fn_out, &ret) < 0) break; }
count++;
} else {
if (un_out) { if (check_sam_write1(un_out, header, b, fn_un_out, &ret) < 0) break; }
}
}
if (r < -1) {
fprintf(stderr, "[main_samview] truncated file.\n");
ret = 1;
}
bam_destroy1(b);
} else { // retrieve alignments in specified regions
int i;
bam1_t *b;
hts_idx_t *idx = sam_index_load(in, fn_in); // load index
if (idx == 0) { // index is unavailable
fprintf(stderr, "[main_samview] random alignment retrieval only works for indexed BAM or CRAM files.\n");
ret = 1;
goto view_end;
}
b = bam_init1();
for (i = optind + 1; i < argc; ++i) {
int result;
hts_itr_t *iter = sam_itr_querys(idx, header, argv[i]); // parse a region in the format like `chr2:100-200'
if (iter == NULL) { // region invalid or reference name not found
int beg, end;
if (hts_parse_reg(argv[i], &beg, &end))
fprintf(stderr, "[main_samview] region \"%s\" specifies an unknown reference name. Continue anyway.\n", argv[i]);
else
fprintf(stderr, "[main_samview] region \"%s\" could not be parsed. Continue anyway.\n", argv[i]);
continue;
}
// fetch alignments
while ((result = sam_itr_next(in, iter, b)) >= 0) {
if (!process_aln(header, b, &settings)) {
if (!is_count) { if (check_sam_write1(out, header, b, fn_out, &ret) < 0) break; }
count++;
} else {
if (un_out) { if (check_sam_write1(un_out, header, b, fn_un_out, &ret) < 0) break; }
}
}
hts_itr_destroy(iter);
if (result < -1) {
fprintf(stderr, "[main_samview] retrieval of region \"%s\" failed due to truncated file or corrupt BAM index file\n", argv[i]);
ret = 1;
break;
}
}
bam_destroy1(b);
hts_idx_destroy(idx); // destroy the BAM index
}
view_end:
if (is_count && ret == 0)
printf("%" PRId64 "\n", count);
// close files, free and return
if (in) check_sam_close("view", in, fn_in, "standard input", &ret);
if (out) check_sam_close("view", out, fn_out, "standard output", &ret);
if (un_out) check_sam_close("view", un_out, fn_un_out, "file", &ret);
free(fn_list); free(fn_out); free(settings.library); free(fn_un_out);
sam_global_args_free(&ga);
if ( header ) bam_hdr_destroy(header);
if (settings.bed) bed_destroy(settings.bed);
if (settings.rghash) {
khint_t k;
for (k = 0; k < kh_end(settings.rghash); ++k)
if (kh_exist(settings.rghash, k)) free((char*)kh_key(settings.rghash, k));
kh_destroy(rg, settings.rghash);
}
if (settings.remove_aux_len) {
free(settings.remove_aux);
}
return ret;
}
static int usage(FILE *fp, int exit_status, int is_long_help)
{
fprintf(fp,
"\n"
"Usage: samtools view [options] <in.bam>|<in.sam>|<in.cram> [region ...]\n"
"\n"
"Options:\n"
// output options
" -b output BAM\n"
" -C output CRAM (requires -T)\n"
" -1 use fast BAM compression (implies -b)\n"
" -u uncompressed BAM output (implies -b)\n"
" -h include header in SAM output\n"
" -H print SAM header only (no alignments)\n"
" -c print only the count of matching records\n"
" -o FILE output file name [stdout]\n"
" -U FILE output reads not selected by filters to FILE [null]\n"
// extra input
" -t FILE FILE listing reference names and lengths (see long help) [null]\n"
// read filters
" -L FILE only include reads overlapping this BED FILE [null]\n"
" -r STR only include reads in read group STR [null]\n"
" -R FILE only include reads with read group listed in FILE [null]\n"
" -q INT only include reads with mapping quality >= INT [0]\n"
" -l STR only include reads in library STR [null]\n"
" -m INT only include reads with number of CIGAR operations consuming\n"
" query sequence >= INT [0]\n"
" -f INT only include reads with all bits set in INT set in FLAG [0]\n"
" -F INT only include reads with none of the bits set in INT set in FLAG [0]\n"
// read processing
" -x STR read tag to strip (repeatable) [null]\n"
" -B collapse the backward CIGAR operation\n"
" -s FLOAT integer part sets seed of random number generator [0];\n"
" rest sets fraction of templates to subsample [no subsampling]\n"
// general options
" -@, --threads INT\n"
" number of BAM/CRAM compression threads [0]\n"
" -? print long help, including note about region specification\n"
" -S ignored (input format is auto-detected)\n");
sam_global_opt_help(fp, "-.O.T");
fprintf(fp, "\n");
if (is_long_help)
fprintf(fp,
"Notes:\n"
"\n"
"1. This command now auto-detects the input format (BAM/CRAM/SAM).\n"
" Further control over the CRAM format can be specified by using the\n"
" --output-fmt-option, e.g. to specify the number of sequences per slice\n"
" and to use avoid reference based compression:\n"
"\n"
"\tsamtools view -C --output-fmt-option seqs_per_slice=5000 \\\n"
"\t --output-fmt-option no_ref -o out.cram in.bam\n"
"\n"
" Options can also be specified as a comma separated list within the\n"
" --output-fmt value too. For example this is equivalent to the above\n"
"\n"
"\tsamtools view --output-fmt cram,seqs_per_slice=5000,no_ref \\\n"
"\t -o out.cram in.bam\n"
"\n"
"2. The file supplied with `-t' is SPACE/TAB delimited with the first\n"
" two fields of each line consisting of the reference name and the\n"
" corresponding sequence length. The `.fai' file generated by \n"
" `samtools faidx' is suitable for use as this file. This may be an\n"
" empty file if reads are unaligned.\n"
"\n"
"3. SAM->BAM conversion: samtools view -bT ref.fa in.sam.gz\n"
"\n"
"4. BAM->SAM conversion: samtools view -h in.bam\n"
"\n"
"5. A region should be presented in one of the following formats:\n"
" `chr1', `chr2:1,000' and `chr3:1000-2,000'. When a region is\n"
" specified, the input alignment file must be a sorted and indexed\n"
" alignment (BAM/CRAM) file.\n"
"\n"
"6. Option `-u' is preferred over `-b' when the output is piped to\n"
" another samtools command.\n"
"\n");
return exit_status;
}
/*
* Performs pileup
* @param conf configuration for this pileup
* @param n number of files specified in fn
* @param fn filenames
*/
static int mpileup(mplp_conf_t *conf, int n, char **fn)
{
extern void *bcf_call_add_rg(void *rghash, const char *hdtext, const char *list);
extern void bcf_call_del_rghash(void *rghash);
mplp_aux_t **data;
int i, tid, pos, *n_plp, tid0 = -1, beg0 = 0, end0 = 1u<<29, ref_len, ref_tid = -1, max_depth, max_indel_depth;
const bam_pileup1_t **plp;
bam_mplp_t iter;
bam_hdr_t *h = NULL; /* header of first file in input list */
char *ref;
void *rghash = NULL;
FILE *pileup_fp = NULL;
bcf_callaux_t *bca = NULL;
bcf_callret1_t *bcr = NULL;
bcf_call_t bc;
htsFile *bcf_fp = NULL;
bcf_hdr_t *bcf_hdr = NULL;
bam_sample_t *sm = NULL;
kstring_t buf;
mplp_pileup_t gplp;
memset(&gplp, 0, sizeof(mplp_pileup_t));
memset(&buf, 0, sizeof(kstring_t));
memset(&bc, 0, sizeof(bcf_call_t));
data = calloc(n, sizeof(mplp_aux_t*));
plp = calloc(n, sizeof(bam_pileup1_t*));
n_plp = calloc(n, sizeof(int));
sm = bam_smpl_init();
if (n == 0) {
fprintf(stderr,"[%s] no input file/data given\n", __func__);
exit(1);
}
// read the header of each file in the list and initialize data
for (i = 0; i < n; ++i) {
bam_hdr_t *h_tmp;
data[i] = calloc(1, sizeof(mplp_aux_t));
data[i]->fp = sam_open(fn[i], "rb");
if ( !data[i]->fp )
{
fprintf(stderr, "[%s] failed to open %s: %s\n", __func__, fn[i], strerror(errno));
exit(1);
}
hts_set_fai_filename(data[i]->fp, conf->fai_fname);
data[i]->conf = conf;
h_tmp = sam_hdr_read(data[i]->fp);
if ( !h_tmp ) {
fprintf(stderr,"[%s] fail to read the header of %s\n", __func__, fn[i]);
exit(1);
}
data[i]->h = i? h : h_tmp; // for i==0, "h" has not been set yet
bam_smpl_add(sm, fn[i], (conf->flag&MPLP_IGNORE_RG)? 0 : h_tmp->text);
// Collect read group IDs with PL (platform) listed in pl_list (note: fragile, strstr search)
rghash = bcf_call_add_rg(rghash, h_tmp->text, conf->pl_list);
if (conf->reg) {
hts_idx_t *idx = sam_index_load(data[i]->fp, fn[i]);
if (idx == 0) {
fprintf(stderr, "[%s] fail to load index for %s\n", __func__, fn[i]);
exit(1);
}
if ( (data[i]->iter=sam_itr_querys(idx, data[i]->h, conf->reg)) == 0) {
fprintf(stderr, "[E::%s] fail to parse region '%s'\n", __func__, conf->reg);
exit(1);
}
if (i == 0) tid0 = data[i]->iter->tid, beg0 = data[i]->iter->beg, end0 = data[i]->iter->end;
hts_idx_destroy(idx);
}
if (i == 0) h = h_tmp; /* save the header of first file in list */
else {
// FIXME: to check consistency
bam_hdr_destroy(h_tmp);
}
}
// allocate data storage proportionate to number of samples being studied sm->n
gplp.n = sm->n;
gplp.n_plp = calloc(sm->n, sizeof(int));
gplp.m_plp = calloc(sm->n, sizeof(int));
gplp.plp = calloc(sm->n, sizeof(bam_pileup1_t*));
fprintf(stderr, "[%s] %d samples in %d input files\n", __func__, sm->n, n);
// write the VCF header
if (conf->flag & MPLP_BCF)
{
const char *mode;
if ( conf->flag & MPLP_VCF )
mode = (conf->flag&MPLP_NO_COMP)? "wu" : "wz"; // uncompressed VCF or compressed VCF
else
mode = (conf->flag&MPLP_NO_COMP)? "wub" : "wb"; // uncompressed BCF or compressed BCF
bcf_fp = bcf_open(conf->output_fname? conf->output_fname : "-", mode);
if (bcf_fp == NULL) {
fprintf(stderr, "[%s] failed to write to %s: %s\n", __func__, conf->output_fname? conf->output_fname : "standard output", strerror(errno));
exit(1);
}
bcf_hdr = bcf_hdr_init("w");
kstring_t str = {0,0,0};
ksprintf(&str, "##samtoolsVersion=%s+htslib-%s\n",samtools_version(),hts_version());
bcf_hdr_append(bcf_hdr, str.s);
str.l = 0;
ksprintf(&str, "##samtoolsCommand=samtools mpileup");
for (i=1; i<conf->argc; i++) ksprintf(&str, " %s", conf->argv[i]);
kputc('\n', &str);
bcf_hdr_append(bcf_hdr, str.s);
if (conf->fai_fname)
{
str.l = 0;
ksprintf(&str, "##reference=file://%s\n", conf->fai_fname);
bcf_hdr_append(bcf_hdr, str.s);
}
// todo: use/write new BAM header manipulation routines, fill also UR, M5
for (i=0; i<h->n_targets; i++)
{
str.l = 0;
ksprintf(&str, "##contig=<ID=%s,length=%d>", h->target_name[i], h->target_len[i]);
bcf_hdr_append(bcf_hdr, str.s);
}
free(str.s);
bcf_hdr_append(bcf_hdr,"##ALT=<ID=X,Description=\"Represents allele(s) other than observed.\">");
bcf_hdr_append(bcf_hdr,"##INFO=<ID=INDEL,Number=0,Type=Flag,Description=\"Indicates that the variant is an INDEL.\">");
bcf_hdr_append(bcf_hdr,"##INFO=<ID=IDV,Number=1,Type=Integer,Description=\"Maximum number of reads supporting an indel\">");
bcf_hdr_append(bcf_hdr,"##INFO=<ID=IMF,Number=1,Type=Float,Description=\"Maximum fraction of reads supporting an indel\">");
bcf_hdr_append(bcf_hdr,"##INFO=<ID=DP,Number=1,Type=Integer,Description=\"Raw read depth\">");
bcf_hdr_append(bcf_hdr,"##INFO=<ID=VDB,Number=1,Type=Float,Description=\"Variant Distance Bias for filtering splice-site artefacts in RNA-seq data (bigger is better)\",Version=\"3\">");
bcf_hdr_append(bcf_hdr,"##INFO=<ID=RPB,Number=1,Type=Float,Description=\"Mann-Whitney U test of Read Position Bias (bigger is better)\">");
bcf_hdr_append(bcf_hdr,"##INFO=<ID=MQB,Number=1,Type=Float,Description=\"Mann-Whitney U test of Mapping Quality Bias (bigger is better)\">");
bcf_hdr_append(bcf_hdr,"##INFO=<ID=BQB,Number=1,Type=Float,Description=\"Mann-Whitney U test of Base Quality Bias (bigger is better)\">");
bcf_hdr_append(bcf_hdr,"##INFO=<ID=MQSB,Number=1,Type=Float,Description=\"Mann-Whitney U test of Mapping Quality vs Strand Bias (bigger is better)\">");
#if CDF_MWU_TESTS
bcf_hdr_append(bcf_hdr,"##INFO=<ID=RPB2,Number=1,Type=Float,Description=\"Mann-Whitney U test of Read Position Bias [CDF] (bigger is better)\">");
bcf_hdr_append(bcf_hdr,"##INFO=<ID=MQB2,Number=1,Type=Float,Description=\"Mann-Whitney U test of Mapping Quality Bias [CDF] (bigger is better)\">");
bcf_hdr_append(bcf_hdr,"##INFO=<ID=BQB2,Number=1,Type=Float,Description=\"Mann-Whitney U test of Base Quality Bias [CDF] (bigger is better)\">");
bcf_hdr_append(bcf_hdr,"##INFO=<ID=MQSB2,Number=1,Type=Float,Description=\"Mann-Whitney U test of Mapping Quality vs Strand Bias [CDF] (bigger is better)\">");
#endif
bcf_hdr_append(bcf_hdr,"##INFO=<ID=SGB,Number=1,Type=Float,Description=\"Segregation based metric.\">");
bcf_hdr_append(bcf_hdr,"##INFO=<ID=MQ0F,Number=1,Type=Float,Description=\"Fraction of MQ0 reads (smaller is better)\">");
bcf_hdr_append(bcf_hdr,"##INFO=<ID=I16,Number=16,Type=Float,Description=\"Auxiliary tag used for calling, see description of bcf_callret1_t in bam2bcf.h\">");
bcf_hdr_append(bcf_hdr,"##INFO=<ID=QS,Number=R,Type=Float,Description=\"Auxiliary tag used for calling\">");
bcf_hdr_append(bcf_hdr,"##FORMAT=<ID=PL,Number=G,Type=Integer,Description=\"List of Phred-scaled genotype likelihoods\">");
if ( conf->fmt_flag&B2B_FMT_DP )
bcf_hdr_append(bcf_hdr,"##FORMAT=<ID=DP,Number=1,Type=Integer,Description=\"Number of high-quality bases\">");
if ( conf->fmt_flag&B2B_FMT_DV )
bcf_hdr_append(bcf_hdr,"##FORMAT=<ID=DV,Number=1,Type=Integer,Description=\"Number of high-quality non-reference bases\">");
if ( conf->fmt_flag&B2B_FMT_DPR )
bcf_hdr_append(bcf_hdr,"##FORMAT=<ID=DPR,Number=R,Type=Integer,Description=\"Number of high-quality bases observed for each allele\">");
if ( conf->fmt_flag&B2B_INFO_DPR )
bcf_hdr_append(bcf_hdr,"##INFO=<ID=DPR,Number=R,Type=Integer,Description=\"Number of high-quality bases observed for each allele\">");
if ( conf->fmt_flag&B2B_FMT_DP4 )
bcf_hdr_append(bcf_hdr,"##FORMAT=<ID=DP4,Number=4,Type=Integer,Description=\"Number of high-quality ref-fwd, ref-reverse, alt-fwd and alt-reverse bases\">");
if ( conf->fmt_flag&B2B_FMT_SP )
bcf_hdr_append(bcf_hdr,"##FORMAT=<ID=SP,Number=1,Type=Integer,Description=\"Phred-scaled strand bias P-value\">");
for (i=0; i<sm->n; i++)
bcf_hdr_add_sample(bcf_hdr, sm->smpl[i]);
bcf_hdr_add_sample(bcf_hdr, NULL);
bcf_hdr_write(bcf_fp, bcf_hdr);
bca = bcf_call_init(-1., conf->min_baseQ);
bcr = calloc(sm->n, sizeof(bcf_callret1_t));
bca->rghash = rghash;
bca->openQ = conf->openQ, bca->extQ = conf->extQ, bca->tandemQ = conf->tandemQ;
bca->min_frac = conf->min_frac;
bca->min_support = conf->min_support;
bca->per_sample_flt = conf->flag & MPLP_PER_SAMPLE;
bc.bcf_hdr = bcf_hdr;
bc.n = sm->n;
bc.PL = malloc(15 * sm->n * sizeof(*bc.PL));
if (conf->fmt_flag)
{
assert( sizeof(float)==sizeof(int32_t) );
bc.DP4 = malloc(sm->n * sizeof(int32_t) * 4);
bc.fmt_arr = malloc(sm->n * sizeof(float)); // all fmt_flag fields
if ( conf->fmt_flag&(B2B_INFO_DPR|B2B_FMT_DPR) )
{
// first B2B_MAX_ALLELES fields for total numbers, the rest per-sample
bc.DPR = malloc((sm->n+1)*B2B_MAX_ALLELES*sizeof(int32_t));
for (i=0; i<sm->n; i++)
bcr[i].DPR = bc.DPR + (i+1)*B2B_MAX_ALLELES;
}
}
}
else {
pileup_fp = conf->output_fname? fopen(conf->output_fname, "w") : stdout;
if (pileup_fp == NULL) {
fprintf(stderr, "[%s] failed to write to %s: %s\n", __func__, conf->output_fname, strerror(errno));
exit(1);
}
}
if (tid0 >= 0 && conf->fai) { // region is set
ref = faidx_fetch_seq(conf->fai, h->target_name[tid0], 0, 0x7fffffff, &ref_len);
ref_tid = tid0;
for (i = 0; i < n; ++i) data[i]->ref = ref, data[i]->ref_id = tid0;
} else ref_tid = -1, ref = 0;
// begin pileup
iter = bam_mplp_init(n, mplp_func, (void**)data);
if ( conf->flag & MPLP_SMART_OVERLAPS ) bam_mplp_init_overlaps(iter);
max_depth = conf->max_depth;
if (max_depth * sm->n > 1<<20)
fprintf(stderr, "(%s) Max depth is above 1M. Potential memory hog!\n", __func__);
if (max_depth * sm->n < 8000) {
max_depth = 8000 / sm->n;
fprintf(stderr, "<%s> Set max per-file depth to %d\n", __func__, max_depth);
}
max_indel_depth = conf->max_indel_depth * sm->n;
bam_mplp_set_maxcnt(iter, max_depth);
bcf1_t *bcf_rec = bcf_init1();
int ret;
while ( (ret=bam_mplp_auto(iter, &tid, &pos, n_plp, plp)) > 0) {
if (conf->reg && (pos < beg0 || pos >= end0)) continue; // out of the region requested
if (conf->bed && tid >= 0 && !bed_overlap(conf->bed, h->target_name[tid], pos, pos+1)) continue;
if (tid != ref_tid) {
free(ref); ref = 0;
if (conf->fai) ref = faidx_fetch_seq(conf->fai, h->target_name[tid], 0, 0x7fffffff, &ref_len);
for (i = 0; i < n; ++i) data[i]->ref = ref, data[i]->ref_id = tid;
ref_tid = tid;
}
if (conf->flag & MPLP_BCF) {
int total_depth, _ref0, ref16;
for (i = total_depth = 0; i < n; ++i) total_depth += n_plp[i];
group_smpl(&gplp, sm, &buf, n, fn, n_plp, plp, conf->flag & MPLP_IGNORE_RG);
_ref0 = (ref && pos < ref_len)? ref[pos] : 'N';
ref16 = seq_nt16_table[_ref0];
bcf_callaux_clean(bca, &bc);
for (i = 0; i < gplp.n; ++i)
bcf_call_glfgen(gplp.n_plp[i], gplp.plp[i], ref16, bca, bcr + i);
bc.tid = tid; bc.pos = pos;
bcf_call_combine(gplp.n, bcr, bca, ref16, &bc);
bcf_clear1(bcf_rec);
bcf_call2bcf(&bc, bcf_rec, bcr, conf->fmt_flag, 0, 0);
bcf_write1(bcf_fp, bcf_hdr, bcf_rec);
// call indels; todo: subsampling with total_depth>max_indel_depth instead of ignoring?
if (!(conf->flag&MPLP_NO_INDEL) && total_depth < max_indel_depth && bcf_call_gap_prep(gplp.n, gplp.n_plp, gplp.plp, pos, bca, ref, rghash) >= 0)
{
bcf_callaux_clean(bca, &bc);
for (i = 0; i < gplp.n; ++i)
bcf_call_glfgen(gplp.n_plp[i], gplp.plp[i], -1, bca, bcr + i);
if (bcf_call_combine(gplp.n, bcr, bca, -1, &bc) >= 0) {
bcf_clear1(bcf_rec);
bcf_call2bcf(&bc, bcf_rec, bcr, conf->fmt_flag, bca, ref);
bcf_write1(bcf_fp, bcf_hdr, bcf_rec);
}
}
} else {
fprintf(pileup_fp, "%s\t%d\t%c", h->target_name[tid], pos + 1, (ref && pos < ref_len)? ref[pos] : 'N');
for (i = 0; i < n; ++i) {
int j, cnt;
for (j = cnt = 0; j < n_plp[i]; ++j) {
const bam_pileup1_t *p = plp[i] + j;
if (bam_get_qual(p->b)[p->qpos] >= conf->min_baseQ) ++cnt;
}
fprintf(pileup_fp, "\t%d\t", cnt);
if (n_plp[i] == 0) {
fputs("*\t*", pileup_fp);
if (conf->flag & MPLP_PRINT_MAPQ) fputs("\t*", pileup_fp);
if (conf->flag & MPLP_PRINT_POS) fputs("\t*", pileup_fp);
} else {
for (j = 0; j < n_plp[i]; ++j) {
const bam_pileup1_t *p = plp[i] + j;
if (bam_get_qual(p->b)[p->qpos] >= conf->min_baseQ)
pileup_seq(pileup_fp, plp[i] + j, pos, ref_len, ref);
}
putc('\t', pileup_fp);
for (j = 0; j < n_plp[i]; ++j) {
const bam_pileup1_t *p = plp[i] + j;
int c = bam_get_qual(p->b)[p->qpos];
if (c >= conf->min_baseQ) {
c = c + 33 < 126? c + 33 : 126;
putc(c, pileup_fp);
}
}
if (conf->flag & MPLP_PRINT_MAPQ) {
putc('\t', pileup_fp);
for (j = 0; j < n_plp[i]; ++j) {
const bam_pileup1_t *p = plp[i] + j;
int c = bam_get_qual(p->b)[p->qpos];
if ( c < conf->min_baseQ ) continue;
c = plp[i][j].b->core.qual + 33;
if (c > 126) c = 126;
putc(c, pileup_fp);
}
}
if (conf->flag & MPLP_PRINT_POS) {
putc('\t', pileup_fp);
for (j = 0; j < n_plp[i]; ++j) {
if (j > 0) putc(',', pileup_fp);
fprintf(pileup_fp, "%d", plp[i][j].qpos + 1); // FIXME: printf() is very slow...
}
}
}
}
putc('\n', pileup_fp);
}
}
// clean up
free(bc.tmp.s);
bcf_destroy1(bcf_rec);
if (bcf_fp)
{
hts_close(bcf_fp);
bcf_hdr_destroy(bcf_hdr);
bcf_call_destroy(bca);
free(bc.PL);
free(bc.DP4);
free(bc.DPR);
free(bc.fmt_arr);
free(bcr);
}
if (pileup_fp && conf->output_fname) fclose(pileup_fp);
bam_smpl_destroy(sm); free(buf.s);
for (i = 0; i < gplp.n; ++i) free(gplp.plp[i]);
free(gplp.plp); free(gplp.n_plp); free(gplp.m_plp);
bcf_call_del_rghash(rghash);
bam_mplp_destroy(iter);
bam_hdr_destroy(h);
for (i = 0; i < n; ++i) {
sam_close(data[i]->fp);
if (data[i]->iter) hts_itr_destroy(data[i]->iter);
free(data[i]);
}
free(data); free(plp); free(ref); free(n_plp);
return ret;
}
int main_pad2unpad(int argc, char *argv[])
{
samFile *in = 0, *out = 0;
bam_hdr_t *h = 0, *h_fix = 0;
faidx_t *fai = 0;
int c, compress_level = -1, is_long_help = 0;
char in_mode[5], out_mode[6], *fn_out = 0, *fn_list = 0;
int ret=0;
sam_global_args ga = SAM_GLOBAL_ARGS_INIT;
static const struct option lopts[] = {
SAM_OPT_GLOBAL_OPTIONS('-', 0, 0, 0, 'T'),
{ NULL, 0, NULL, 0 }
};
/* parse command-line options */
strcpy(in_mode, "r"); strcpy(out_mode, "w");
while ((c = getopt_long(argc, argv, "SCso:u1T:?", lopts, NULL)) >= 0) {
switch (c) {
case 'S': break;
case 'C': hts_parse_format(&ga.out, "cram"); break;
case 's': assert(compress_level == -1); hts_parse_format(&ga.out, "sam"); break;
case 'o': fn_out = strdup(optarg); break;
case 'u':
compress_level = 0;
if (ga.out.format == unknown_format)
hts_parse_format(&ga.out, "bam");
break;
case '1':
compress_level = 1;
if (ga.out.format == unknown_format)
hts_parse_format(&ga.out, "bam");
break;
case '?': is_long_help = 1; break;
default: if (parse_sam_global_opt(c, optarg, lopts, &ga) == 0) break;
fprintf(stderr, "[bam_fillmd] unrecognized option '-%c'\n\n", c);
return usage(is_long_help);
}
}
if (argc == optind) return usage(is_long_help);
strcat(out_mode, "h");
if (compress_level >= 0) {
char tmp[2];
tmp[0] = compress_level + '0'; tmp[1] = '\0';
strcat(out_mode, tmp);
}
// Load FASTA reference (also needed for SAM -> BAM if missing header)
if (ga.reference) {
fn_list = samfaipath(ga.reference);
fai = fai_load(ga.reference);
}
// open file handlers
if ((in = sam_open_format(argv[optind], in_mode, &ga.in)) == 0) {
fprintf(stderr, "[depad] failed to open \"%s\" for reading.\n", argv[optind]);
ret = 1;
goto depad_end;
}
if (fn_list && hts_set_fai_filename(in, fn_list) != 0) {
fprintf(stderr, "[depad] failed to load reference file \"%s\".\n", fn_list);
ret = 1;
goto depad_end;
}
if ((h = sam_hdr_read(in)) == 0) {
fprintf(stderr, "[depad] failed to read the header from \"%s\".\n", argv[optind]);
ret = 1;
goto depad_end;
}
if (fai) {
h_fix = fix_header(h, fai);
} else {
fprintf(stderr, "[depad] Warning - reference lengths will not be corrected without FASTA reference\n");
h_fix = h;
}
char wmode[2];
strcat(out_mode, sam_open_mode(wmode, fn_out, NULL)==0 ? wmode : "b");
if ((out = sam_open_format(fn_out? fn_out : "-", out_mode, &ga.out)) == 0) {
fprintf(stderr, "[depad] failed to open \"%s\" for writing.\n", fn_out? fn_out : "standard output");
ret = 1;
goto depad_end;
}
// Reference-based CRAM won't work unless we also create a new reference.
// We could embed this, but for now we take the easy option.
if (ga.out.format == cram)
hts_set_opt(out, CRAM_OPT_NO_REF, 1);
if (sam_hdr_write(out, h_fix) != 0) {
fprintf(stderr, "[depad] failed to write header.\n");
ret = 1;
goto depad_end;
}
// Do the depad
ret = bam_pad2unpad(in, out, h, fai);
depad_end:
// close files, free and return
if (fai) fai_destroy(fai);
if (h) bam_hdr_destroy(h);
sam_close(in);
sam_close(out);
free(fn_list); free(fn_out);
return ret;
}
