int bam_mating(int argc, char *argv[])
{
samFile *in, *out;
int c, remove_reads = 0, proper_pair_check = 1, add_ct = 0;
sam_global_args ga = SAM_GLOBAL_ARGS_INIT;
char wmode[3] = {'w', 'b', 0};
static const struct option lopts[] = {
SAM_OPT_GLOBAL_OPTIONS('-', 0, 'O', 0, 0),
{ NULL, 0, NULL, 0 }
};
// parse args
if (argc == 1) { usage(stdout); return 0; }
while ((c = getopt_long(argc, argv, "rpcO:", lopts, NULL)) >= 0) {
switch (c) {
case 'r': remove_reads = 1; break;
case 'p': proper_pair_check = 0; break;
case 'c': add_ct = 1; break;
default: if (parse_sam_global_opt(c, optarg, lopts, &ga) == 0) break;
/* else fall-through */
case '?': usage(stderr); return 1;
}
}
if (optind+1 >= argc) { usage(stderr); return 1; }
// init
if ((in = sam_open_format(argv[optind], "rb", &ga.in)) == NULL) {
fprintf(stderr, "[bam_mating] cannot open input file\n");
return 1;
}
sam_open_mode(wmode+1, argv[optind+1], NULL);
if ((out = sam_open_format(argv[optind+1], wmode, &ga.out)) == NULL) {
fprintf(stderr, "[bam_mating] cannot open output file\n");
return 1;
}
// run
bam_mating_core(in, out, remove_reads, proper_pair_check, add_ct);
// cleanup
sam_close(in); sam_close(out);
sam_global_args_free(&ga);
return 0;
}
static bool init(const parsed_opts_t* opts, state_t** state_out) {
state_t* retval = (state_t*) calloc(1, sizeof(state_t));
if (retval == NULL) {
fprintf(stderr, "[init] Out of memory allocating state struct.\n");
return false;
}
*state_out = retval;
// Open files
retval->input_file = sam_open_format(opts->input_name, "r", &opts->ga.in);
if (retval->input_file == NULL) {
fprintf(stderr, "[init] Could not open input file: %s\n", opts->input_name);
return false;
}
retval->input_header = sam_hdr_read(retval->input_file);
retval->output_header = bam_hdr_dup(retval->input_header);
retval->output_file = sam_open_format(opts->output_name == NULL?"-":opts->output_name, "w", &opts->ga.out);
if (retval->output_file == NULL) {
print_error_errno("addreplacerg", "Could not open output file: %s\n", opts->output_name);
return false;
}
if (opts->rg_line) {
// Append new RG line to header.
// Check does not already exist
if ( confirm_rg(retval->output_header, opts->rg_id) ) {
fprintf(stderr, "[init] ID of new RG line specified conflicts with that of an existing header RG line. Overwrite not yet implemented.\n");
return false;
}
retval->rg_id = strdup(opts->rg_id);
size_t new_len = strlen( retval->output_header->text ) + strlen( opts->rg_line ) + 2;
char* new_header = malloc(new_len);
if (!new_header) {
fprintf(stderr, "[init] Out of memory whilst writing new header.\n");
return false;
}
sprintf(new_header,"%s%s\n", retval->output_header->text, opts->rg_line);
free(retval->output_header->text);
retval->output_header->text = new_header;
retval->output_header->l_text = (int)new_len - 1;
} else {
if (opts->rg_id) {
// Confirm what has been supplied exists
if ( !confirm_rg(retval->output_header, opts->rg_id) ) {
fprintf(stderr, "RG ID supplied does not exist in header. Supply full @RG line with -r instead?\n");
return false;
}
retval->rg_id = strdup(opts->rg_id);
} else {
if ((retval->rg_id = get_first_rgid(retval->output_header)) == NULL ) {
fprintf(stderr, "No RG specified on command line or in existing header.\n");
return false;
}
}
}
switch (opts->mode) {
case overwrite_all:
retval->mode_func = &overwrite_all_func;
break;
case orphan_only:
retval->mode_func = &orphan_only_func;
break;
}
return true;
}
int main_bedcov(int argc, char *argv[])
{
gzFile fp;
kstring_t str;
kstream_t *ks;
hts_idx_t **idx;
aux_t **aux;
int *n_plp, dret, i, n, c, min_mapQ = 0;
int64_t *cnt;
const bam_pileup1_t **plp;
int usage = 0;
sam_global_args ga = SAM_GLOBAL_ARGS_INIT;
static const struct option lopts[] = {
SAM_OPT_GLOBAL_OPTIONS('-', 0, '-', '-', 0),
{ NULL, 0, NULL, 0 }
};
while ((c = getopt_long(argc, argv, "Q:", lopts, NULL)) >= 0) {
switch (c) {
case 'Q': min_mapQ = atoi(optarg); break;
default: if (parse_sam_global_opt(c, optarg, lopts, &ga) == 0) break;
/* else fall-through */
case '?': usage = 1; break;
}
if (usage) break;
}
if (usage || optind + 2 > argc) {
fprintf(pysam_stderr, "Usage: samtools bedcov [options] <in.bed> <in1.bam> [...]\n\n");
fprintf(pysam_stderr, " -Q INT Only count bases of at least INT quality [0]\n");
sam_global_opt_help(pysam_stderr, "-.--.");
return 1;
}
memset(&str, 0, sizeof(kstring_t));
n = argc - optind - 1;
aux = calloc(n, sizeof(aux_t*));
idx = calloc(n, sizeof(hts_idx_t*));
for (i = 0; i < n; ++i) {
aux[i] = calloc(1, sizeof(aux_t));
aux[i]->min_mapQ = min_mapQ;
aux[i]->fp = sam_open_format(argv[i+optind+1], "r", &ga.in);
if (aux[i]->fp)
idx[i] = sam_index_load(aux[i]->fp, argv[i+optind+1]);
if (aux[i]->fp == 0 || idx[i] == 0) {
fprintf(pysam_stderr, "ERROR: fail to open index BAM file '%s'\n", argv[i+optind+1]);
return 2;
}
// TODO bgzf_set_cache_size(aux[i]->fp, 20);
aux[i]->header = sam_hdr_read(aux[i]->fp);
if (aux[i]->header == NULL) {
fprintf(pysam_stderr, "ERROR: failed to read header for '%s'\n",
argv[i+optind+1]);
return 2;
}
}
cnt = calloc(n, 8);
fp = gzopen(argv[optind], "rb");
ks = ks_init(fp);
n_plp = calloc(n, sizeof(int));
plp = calloc(n, sizeof(bam_pileup1_t*));
while (ks_getuntil(ks, KS_SEP_LINE, &str, &dret) >= 0) {
char *p, *q;
int tid, beg, end, pos;
bam_mplp_t mplp;
for (p = q = str.s; *p && *p != '\t'; ++p);
if (*p != '\t') goto bed_error;
*p = 0; tid = bam_name2id(aux[0]->header, q); *p = '\t';
if (tid < 0) goto bed_error;
for (q = p = p + 1; isdigit(*p); ++p);
if (*p != '\t') goto bed_error;
*p = 0; beg = atoi(q); *p = '\t';
for (q = p = p + 1; isdigit(*p); ++p);
if (*p == '\t' || *p == 0) {
int c = *p;
*p = 0; end = atoi(q); *p = c;
} else goto bed_error;
for (i = 0; i < n; ++i) {
if (aux[i]->iter) hts_itr_destroy(aux[i]->iter);
aux[i]->iter = sam_itr_queryi(idx[i], tid, beg, end);
}
mplp = bam_mplp_init(n, read_bam, (void**)aux);
bam_mplp_set_maxcnt(mplp, 64000);
memset(cnt, 0, 8 * n);
while (bam_mplp_auto(mplp, &tid, &pos, n_plp, plp) > 0)
if (pos >= beg && pos < end)
for (i = 0; i < n; ++i) cnt[i] += n_plp[i];
for (i = 0; i < n; ++i) {
kputc('\t', &str);
kputl(cnt[i], &str);
}
fputs(str.s, pysam_stdout) & fputc('\n', pysam_stdout);
bam_mplp_destroy(mplp);
continue;
bed_error:
fprintf(pysam_stderr, "Errors in BED line '%s'\n", str.s);
}
free(n_plp); free(plp);
ks_destroy(ks);
gzclose(fp);
free(cnt);
for (i = 0; i < n; ++i) {
if (aux[i]->iter) hts_itr_destroy(aux[i]->iter);
hts_idx_destroy(idx[i]);
bam_hdr_destroy(aux[i]->header);
sam_close(aux[i]->fp);
free(aux[i]);
}
free(aux); free(idx);
free(str.s);
sam_global_args_free(&ga);
return 0;
}
int bam_fillmd(int argc, char *argv[])
{
int c, flt_flag, tid = -2, ret, len, is_bam_out, is_uncompressed, max_nm, is_realn, capQ, baq_flag;
samFile *fp = NULL, *fpout = NULL;
bam_hdr_t *header = NULL;
faidx_t *fai = NULL;
char *ref = NULL, mode_w[8], *ref_file;
bam1_t *b = NULL;
sam_global_args ga = SAM_GLOBAL_ARGS_INIT;
static const struct option lopts[] = {
SAM_OPT_GLOBAL_OPTIONS('-', 0, 0, 0, 0),
{ NULL, 0, NULL, 0 }
};
flt_flag = UPDATE_NM | UPDATE_MD;
is_bam_out = is_uncompressed = is_realn = max_nm = capQ = baq_flag = 0;
strcpy(mode_w, "w");
while ((c = getopt_long(argc, argv, "EqreuNhbSC:n:Ad", lopts, NULL)) >= 0) {
switch (c) {
case 'r':
is_realn = 1;
break;
case 'e':
flt_flag |= USE_EQUAL;
break;
case 'd':
flt_flag |= DROP_TAG;
break;
case 'q':
flt_flag |= BIN_QUAL;
break;
case 'h':
flt_flag |= HASH_QNM;
break;
case 'N':
flt_flag &= ~(UPDATE_MD|UPDATE_NM);
break;
case 'b':
is_bam_out = 1;
break;
case 'u':
is_uncompressed = is_bam_out = 1;
break;
case 'S':
break;
case 'n':
max_nm = atoi(optarg);
break;
case 'C':
capQ = atoi(optarg);
break;
case 'A':
baq_flag |= 1;
break;
case 'E':
baq_flag |= 2;
break;
default:
if (parse_sam_global_opt(c, optarg, lopts, &ga) == 0) break;
fprintf(stderr, "[bam_fillmd] unrecognized option '-%c'\n\n", c);
/* else fall-through */
case '?':
return calmd_usage();
}
}
if (is_bam_out) strcat(mode_w, "b");
else strcat(mode_w, "h");
if (is_uncompressed) strcat(mode_w, "0");
if (optind + (ga.reference == NULL) >= argc)
return calmd_usage();
fp = sam_open_format(argv[optind], "r", &ga.in);
if (fp == NULL) {
print_error_errno("calmd", "Failed to open input file '%s'", argv[optind]);
return 1;
}
header = sam_hdr_read(fp);
if (header == NULL || header->n_targets == 0) {
fprintf(stderr, "[bam_fillmd] input SAM does not have header. Abort!\n");
goto fail;
}
fpout = sam_open_format("-", mode_w, &ga.out);
if (fpout == NULL) {
print_error_errno("calmd", "Failed to open output");
goto fail;
}
if (sam_hdr_write(fpout, header) < 0) {
print_error_errno("calmd", "Failed to write sam header");
goto fail;
}
ref_file = argc > optind + 1 ? argv[optind+1] : ga.reference;
fai = fai_load(ref_file);
if (!fai) {
print_error_errno("calmd", "Failed to open reference file '%s'", ref_file);
goto fail;
}
b = bam_init1();
if (!b) {
fprintf(stderr, "[bam_fillmd] Failed to allocate bam struct\n");
goto fail;
}
while ((ret = sam_read1(fp, header, b)) >= 0) {
if (b->core.tid >= 0) {
if (tid != b->core.tid) {
free(ref);
ref = fai_fetch(fai, header->target_name[b->core.tid], &len);
tid = b->core.tid;
if (ref == 0) { // FIXME: Should this always be fatal?
fprintf(stderr, "[bam_fillmd] fail to find sequence '%s' in the reference.\n",
header->target_name[tid]);
if (is_realn || capQ > 10) goto fail; // Would otherwise crash
}
}
if (is_realn) sam_prob_realn(b, ref, len, baq_flag);
if (capQ > 10) {
int q = sam_cap_mapq(b, ref, len, capQ);
if (b->core.qual > q) b->core.qual = q;
}
if (ref) bam_fillmd1_core(b, ref, len, flt_flag, max_nm);
}
if (sam_write1(fpout, header, b) < 0) {
print_error_errno("calmd", "failed to write to output file");
goto fail;
}
}
if (ret < -1) {
fprintf(stderr, "[bam_fillmd] Error reading input.\n");
goto fail;
}
bam_destroy1(b);
bam_hdr_destroy(header);
free(ref);
fai_destroy(fai);
sam_close(fp);
if (sam_close(fpout) < 0) {
fprintf(stderr, "[bam_fillmd] error when closing output file\n");
return 1;
}
return 0;
fail:
free(ref);
if (b) bam_destroy1(b);
if (header) bam_hdr_destroy(header);
if (fai) fai_destroy(fai);
if (fp) sam_close(fp);
if (fpout) sam_close(fpout);
return 1;
}
int main_depth(int argc, char *argv[])
{
int i, n, tid, reg_tid, beg, end, pos, *n_plp, baseQ = 0, mapQ = 0, min_len = 0;
int all = 0, status = EXIT_SUCCESS, nfiles, max_depth = -1;
const bam_pileup1_t **plp;
char *reg = 0; // specified region
void *bed = 0; // BED data structure
char *file_list = NULL, **fn = NULL;
bam_hdr_t *h = NULL; // BAM header of the 1st input
aux_t **data;
bam_mplp_t mplp;
int last_pos = -1, last_tid = -1, ret;
sam_global_args ga = SAM_GLOBAL_ARGS_INIT;
static const struct option lopts[] = {
SAM_OPT_GLOBAL_OPTIONS('-', 0, '-', '-', 0),
{ NULL, 0, NULL, 0 }
};
// parse the command line
while ((n = getopt_long(argc, argv, "r:b:q:Q:l:f:am:d:", lopts, NULL)) >= 0) {
switch (n) {
case 'l':
min_len = atoi(optarg);
break; // minimum query length
case 'r':
reg = strdup(optarg);
break; // parsing a region requires a BAM header
case 'b':
bed = bed_read(optarg); // BED or position list file can be parsed now
if (!bed) {
print_error_errno("depth", "Could not read file \"%s\"", optarg);
return 1;
}
break;
case 'q':
baseQ = atoi(optarg);
break; // base quality threshold
case 'Q':
mapQ = atoi(optarg);
break; // mapping quality threshold
case 'f':
file_list = optarg;
break;
case 'a':
all++;
break;
case 'd':
case 'm':
max_depth = atoi(optarg);
break; // maximum coverage depth
default:
if (parse_sam_global_opt(n, optarg, lopts, &ga) == 0) break;
/* else fall-through */
case '?':
return usage();
}
}
if (optind == argc && !file_list)
return usage();
// initialize the auxiliary data structures
if (file_list)
{
if ( read_file_list(file_list,&nfiles,&fn) ) return 1;
n = nfiles;
argv = fn;
optind = 0;
}
else
n = argc - optind; // the number of BAMs on the command line
data = calloc(n, sizeof(aux_t*)); // data[i] for the i-th input
reg_tid = 0;
beg = 0;
end = INT_MAX; // set the default region
for (i = 0; i < n; ++i) {
int rf;
data[i] = calloc(1, sizeof(aux_t));
data[i]->fp = sam_open_format(argv[optind+i], "r", &ga.in); // open BAM
if (data[i]->fp == NULL) {
print_error_errno("depth", "Could not open \"%s\"", argv[optind+i]);
status = EXIT_FAILURE;
goto depth_end;
}
rf = SAM_FLAG | SAM_RNAME | SAM_POS | SAM_MAPQ | SAM_CIGAR | SAM_SEQ;
if (baseQ) rf |= SAM_QUAL;
if (hts_set_opt(data[i]->fp, CRAM_OPT_REQUIRED_FIELDS, rf)) {
fprintf(stderr, "Failed to set CRAM_OPT_REQUIRED_FIELDS value\n");
return 1;
}
if (hts_set_opt(data[i]->fp, CRAM_OPT_DECODE_MD, 0)) {
fprintf(stderr, "Failed to set CRAM_OPT_DECODE_MD value\n");
return 1;
}
data[i]->min_mapQ = mapQ; // set the mapQ filter
data[i]->min_len = min_len; // set the qlen filter
data[i]->hdr = sam_hdr_read(data[i]->fp); // read the BAM header
if (data[i]->hdr == NULL) {
fprintf(stderr, "Couldn't read header for \"%s\"\n",
argv[optind+i]);
status = EXIT_FAILURE;
goto depth_end;
}
if (reg) { // if a region is specified
hts_idx_t *idx = sam_index_load(data[i]->fp, argv[optind+i]); // load the index
if (idx == NULL) {
print_error("depth", "can't load index for \"%s\"", argv[optind+i]);
status = EXIT_FAILURE;
goto depth_end;
}
data[i]->iter = sam_itr_querys(idx, data[i]->hdr, reg); // set the iterator
hts_idx_destroy(idx); // the index is not needed any more; free the memory
if (data[i]->iter == NULL) {
print_error("depth", "can't parse region \"%s\"", reg);
status = EXIT_FAILURE;
goto depth_end;
}
}
}
h = data[0]->hdr; // easy access to the header of the 1st BAM
if (reg) {
beg = data[0]->iter->beg; // and to the parsed region coordinates
end = data[0]->iter->end;
reg_tid = data[0]->iter->tid;
}
// the core multi-pileup loop
mplp = bam_mplp_init(n, read_bam, (void**)data); // initialization
if (0 < max_depth)
bam_mplp_set_maxcnt(mplp,max_depth); // set maximum coverage depth
n_plp = calloc(n, sizeof(int)); // n_plp[i] is the number of covering reads from the i-th BAM
plp = calloc(n, sizeof(bam_pileup1_t*)); // plp[i] points to the array of covering reads (internal in mplp)
while ((ret=bam_mplp_auto(mplp, &tid, &pos, n_plp, plp)) > 0) { // come to the next covered position
if (pos < beg || pos >= end) continue; // out of range; skip
if (tid >= h->n_targets) continue; // diff number of @SQ lines per file?
if (all) {
while (tid > last_tid) {
if (last_tid >= 0 && !reg) {
// Deal with remainder or entirety of last tid.
while (++last_pos < h->target_len[last_tid]) {
// Horribly inefficient, but the bed API is an obfuscated black box.
if (bed && bed_overlap(bed, h->target_name[last_tid], last_pos, last_pos + 1) == 0)
continue;
fputs(h->target_name[last_tid], stdout);
printf("\t%d", last_pos+1);
for (i = 0; i < n; i++)
putchar('\t'), putchar('0');
putchar('\n');
}
}
last_tid++;
last_pos = -1;
if (all < 2)
break;
}
// Deal with missing portion of current tid
while (++last_pos < pos) {
if (last_pos < beg) continue; // out of range; skip
if (bed && bed_overlap(bed, h->target_name[tid], last_pos, last_pos + 1) == 0)
continue;
fputs(h->target_name[tid], stdout);
printf("\t%d", last_pos+1);
for (i = 0; i < n; i++)
putchar('\t'), putchar('0');
putchar('\n');
}
last_tid = tid;
last_pos = pos;
}
if (bed && bed_overlap(bed, h->target_name[tid], pos, pos + 1) == 0) continue;
fputs(h->target_name[tid], stdout);
printf("\t%d", pos+1); // a customized printf() would be faster
for (i = 0; i < n; ++i) { // base level filters have to go here
int j, m = 0;
for (j = 0; j < n_plp[i]; ++j) {
const bam_pileup1_t *p = plp[i] + j; // DON'T modfity plp[][] unless you really know
if (p->is_del || p->is_refskip) ++m; // having dels or refskips at tid:pos
else if (bam_get_qual(p->b)[p->qpos] < baseQ) ++m; // low base quality
}
printf("\t%d", n_plp[i] - m); // this the depth to output
}
putchar('\n');
}
if (ret < 0) status = EXIT_FAILURE;
free(n_plp);
free(plp);
bam_mplp_destroy(mplp);
if (all) {
// Handle terminating region
if (last_tid < 0 && reg && all > 1) {
last_tid = reg_tid;
last_pos = beg-1;
}
while (last_tid >= 0 && last_tid < h->n_targets) {
while (++last_pos < h->target_len[last_tid]) {
if (last_pos >= end) break;
if (bed && bed_overlap(bed, h->target_name[last_tid], last_pos, last_pos + 1) == 0)
continue;
fputs(h->target_name[last_tid], stdout);
printf("\t%d", last_pos+1);
for (i = 0; i < n; i++)
putchar('\t'), putchar('0');
putchar('\n');
}
last_tid++;
last_pos = -1;
if (all < 2 || reg)
break;
}
}
depth_end:
for (i = 0; i < n && data[i]; ++i) {
bam_hdr_destroy(data[i]->hdr);
if (data[i]->fp) sam_close(data[i]->fp);
hts_itr_destroy(data[i]->iter);
free(data[i]);
}
free(data);
free(reg);
if (bed) bed_destroy(bed);
if ( file_list )
{
for (i=0; i<n; i++) free(fn[i]);
free(fn);
}
sam_global_args_free(&ga);
return status;
}
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;
}
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;
}
int bam_flagstat(int argc, char *argv[])
{
samFile *fp;
bam_hdr_t *header;
bam_flagstat_t *s;
char b0[16], b1[16];
int c;
enum {
INPUT_FMT_OPTION = CHAR_MAX+1,
};
sam_global_args ga = SAM_GLOBAL_ARGS_INIT;
static const struct option lopts[] = {
SAM_OPT_GLOBAL_OPTIONS('-', 0, '-', '-', '-', '@'),
{NULL, 0, NULL, 0}
};
while ((c = getopt_long(argc, argv, "@:", lopts, NULL)) >= 0) {
switch (c) {
default: if (parse_sam_global_opt(c, optarg, lopts, &ga) == 0) break;
/* else fall-through */
case '?':
usage_exit(samtools_stderr, EXIT_FAILURE);
}
}
if (argc != optind+1) {
if (argc == optind) usage_exit(samtools_stdout, EXIT_SUCCESS);
else usage_exit(samtools_stderr, EXIT_FAILURE);
}
fp = sam_open_format(argv[optind], "r", &ga.in);
if (fp == NULL) {
print_error_errno("flagstat", "Cannot open input file \"%s\"", argv[optind]);
return 1;
}
if (ga.nthreads > 0)
hts_set_threads(fp, ga.nthreads);
if (hts_set_opt(fp, CRAM_OPT_REQUIRED_FIELDS,
SAM_FLAG | SAM_MAPQ | SAM_RNEXT)) {
fprintf(samtools_stderr, "Failed to set CRAM_OPT_REQUIRED_FIELDS value\n");
return 1;
}
if (hts_set_opt(fp, CRAM_OPT_DECODE_MD, 0)) {
fprintf(samtools_stderr, "Failed to set CRAM_OPT_DECODE_MD value\n");
return 1;
}
header = sam_hdr_read(fp);
if (header == NULL) {
fprintf(samtools_stderr, "Failed to read header for \"%s\"\n", argv[optind]);
return 1;
}
s = bam_flagstat_core(fp, header);
fprintf(samtools_stdout, "%lld + %lld in total (QC-passed reads + QC-failed reads)\n", s->n_reads[0], s->n_reads[1]);
fprintf(samtools_stdout, "%lld + %lld secondary\n", s->n_secondary[0], s->n_secondary[1]);
fprintf(samtools_stdout, "%lld + %lld supplementary\n", s->n_supp[0], s->n_supp[1]);
fprintf(samtools_stdout, "%lld + %lld duplicates\n", s->n_dup[0], s->n_dup[1]);
fprintf(samtools_stdout, "%lld + %lld mapped (%s : %s)\n", s->n_mapped[0], s->n_mapped[1], percent(b0, s->n_mapped[0], s->n_reads[0]), percent(b1, s->n_mapped[1], s->n_reads[1]));
fprintf(samtools_stdout, "%lld + %lld paired in sequencing\n", s->n_pair_all[0], s->n_pair_all[1]);
fprintf(samtools_stdout, "%lld + %lld read1\n", s->n_read1[0], s->n_read1[1]);
fprintf(samtools_stdout, "%lld + %lld read2\n", s->n_read2[0], s->n_read2[1]);
fprintf(samtools_stdout, "%lld + %lld properly paired (%s : %s)\n", s->n_pair_good[0], s->n_pair_good[1], percent(b0, s->n_pair_good[0], s->n_pair_all[0]), percent(b1, s->n_pair_good[1], s->n_pair_all[1]));
fprintf(samtools_stdout, "%lld + %lld with itself and mate mapped\n", s->n_pair_map[0], s->n_pair_map[1]);
fprintf(samtools_stdout, "%lld + %lld singletons (%s : %s)\n", s->n_sgltn[0], s->n_sgltn[1], percent(b0, s->n_sgltn[0], s->n_pair_all[0]), percent(b1, s->n_sgltn[1], s->n_pair_all[1]));
fprintf(samtools_stdout, "%lld + %lld with mate mapped to a different chr\n", s->n_diffchr[0], s->n_diffchr[1]);
fprintf(samtools_stdout, "%lld + %lld with mate mapped to a different chr (mapQ>=5)\n", s->n_diffhigh[0], s->n_diffhigh[1]);
free(s);
bam_hdr_destroy(header);
sam_close(fp);
sam_global_args_free(&ga);
return 0;
}
/*
* 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, beg0 = 0, end0 = INT_MAX, ref_len, max_depth, max_indel_depth;
const bam_pileup1_t **plp;
mplp_ref_t mp_ref = MPLP_REF_INIT;
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(pysam_stderr,"[%s] no input file/data given\n", __func__);
exit(EXIT_FAILURE);
}
// 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_format(fn[i], "rb", &conf->ga.in);
if ( !data[i]->fp )
{
fprintf(pysam_stderr, "[%s] failed to open %s: %s\n", __func__, fn[i], strerror(errno));
exit(EXIT_FAILURE);
}
if (hts_set_opt(data[i]->fp, CRAM_OPT_DECODE_MD, 0)) {
fprintf(pysam_stderr, "Failed to set CRAM_OPT_DECODE_MD value\n");
exit(EXIT_FAILURE);
}
if (conf->fai_fname && hts_set_fai_filename(data[i]->fp, conf->fai_fname) != 0) {
fprintf(pysam_stderr, "[%s] failed to process %s: %s\n",
__func__, conf->fai_fname, strerror(errno));
exit(EXIT_FAILURE);
}
data[i]->conf = conf;
data[i]->ref = &mp_ref;
h_tmp = sam_hdr_read(data[i]->fp);
if ( !h_tmp ) {
fprintf(pysam_stderr,"[%s] fail to read the header of %s\n", __func__, fn[i]);
exit(EXIT_FAILURE);
}
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 == NULL) {
fprintf(pysam_stderr, "[%s] fail to load index for %s\n", __func__, fn[i]);
exit(EXIT_FAILURE);
}
if ( (data[i]->iter=sam_itr_querys(idx, h_tmp, conf->reg)) == 0) {
fprintf(pysam_stderr, "[E::%s] fail to parse region '%s' with %s\n", __func__, conf->reg, fn[i]);
exit(EXIT_FAILURE);
}
if (i == 0) beg0 = data[i]->iter->beg, end0 = data[i]->iter->end;
hts_idx_destroy(idx);
}
else
data[i]->iter = NULL;
if (i == 0) h = data[i]->h = h_tmp; // save the header of the first file
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
data[i]->h = h;
}
}
// 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(pysam_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(pysam_stderr, "[%s] failed to write to %s: %s\n", __func__, conf->output_fname? conf->output_fname : "standard output", strerror(errno));
exit(EXIT_FAILURE);
}
// BCF header creation
bcf_hdr = bcf_hdr_init("w");
kstring_t str = {0,0,NULL};
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);
}
// Translate BAM @SQ tags to BCF ##contig tags
// 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=*,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\">");
if ( conf->fmt_flag&B2B_FMT_AD )
bcf_hdr_append(bcf_hdr,"##FORMAT=<ID=AD,Number=R,Type=Integer,Description=\"Allelic depths\">");
if ( conf->fmt_flag&B2B_FMT_ADF )
bcf_hdr_append(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(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(bcf_hdr,"##INFO=<ID=AD,Number=R,Type=Integer,Description=\"Total allelic depths\">");
if ( conf->fmt_flag&B2B_INFO_ADF )
bcf_hdr_append(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(bcf_hdr,"##INFO=<ID=ADR,Number=R,Type=Integer,Description=\"Total allelic depths on the reverse strand\">");
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);
// End of BCF header creation
// Initialise the calling algorithm
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|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
bc.ADR = (int32_t*) malloc((sm->n+1)*B2B_MAX_ALLELES*sizeof(int32_t));
bc.ADF = (int32_t*) malloc((sm->n+1)*B2B_MAX_ALLELES*sizeof(int32_t));
for (i=0; i<sm->n; i++)
{
bcr[i].ADR = bc.ADR + (i+1)*B2B_MAX_ALLELES;
bcr[i].ADF = bc.ADF + (i+1)*B2B_MAX_ALLELES;
}
}
}
}
else {
pileup_fp = conf->output_fname? fopen(conf->output_fname, "w") : pysam_stdout;
if (pileup_fp == NULL) {
fprintf(pysam_stderr, "[%s] failed to write to %s: %s\n", __func__, conf->output_fname, strerror(errno));
exit(EXIT_FAILURE);
}
}
// init 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(pysam_stderr, "(%s) Max depth is above 1M. Potential memory hog!\n", __func__);
if (max_depth * sm->n < 8000) {
max_depth = 8000 / sm->n;
fprintf(pysam_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;
// begin pileup
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;
mplp_get_ref(data[0], tid, &ref, &ref_len);
//printf("tid=%d len=%d ref=%p/%s\n", tid, ref_len, ref, ref);
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;
int c = p->qpos < p->b->core.l_qseq
? bam_get_qual(p->b)[p->qpos]
: 0;
if (c >= 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;
int c = p->qpos < p->b->core.l_qseq
? bam_get_qual(p->b)[p->qpos]
: 0;
if (c >= 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 = p->qpos < p->b->core.l_qseq
? bam_get_qual(p->b)[p->qpos]
: 0;
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);
int last = 0;
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;
if (last++) putc(',', pileup_fp);
fprintf(pileup_fp, "%d", plp[i][j].qpos + 1); // FIXME: fprintf(pysam_stdout, ) 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.ADR);
free(bc.ADF);
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(n_plp);
free(mp_ref.ref[0]);
free(mp_ref.ref[1]);
return ret;
}
int bam_mating(int argc, char *argv[])
{
htsThreadPool p = {NULL, 0};
samFile *in = NULL, *out = NULL;
int c, remove_reads = 0, proper_pair_check = 1, add_ct = 0, res = 1, mate_score = 0;
sam_global_args ga = SAM_GLOBAL_ARGS_INIT;
char wmode[3] = {'w', 'b', 0};
static const struct option lopts[] = {
SAM_OPT_GLOBAL_OPTIONS('-', 0, 'O', 0, 0, '@'),
{ NULL, 0, NULL, 0 }
};
// parse args
if (argc == 1) { usage(stdout); return 0; }
while ((c = getopt_long(argc, argv, "rpcmO:@:", lopts, NULL)) >= 0) {
switch (c) {
case 'r': remove_reads = 1; break;
case 'p': proper_pair_check = 0; break;
case 'c': add_ct = 1; break;
case 'm': mate_score = 1; break;
default: if (parse_sam_global_opt(c, optarg, lopts, &ga) == 0) break;
/* else fall-through */
case '?': usage(stderr); goto fail;
}
}
if (optind+1 >= argc) { usage(stderr); goto fail; }
// init
if ((in = sam_open_format(argv[optind], "rb", &ga.in)) == NULL) {
print_error_errno("fixmate", "cannot open input file");
goto fail;
}
sam_open_mode(wmode+1, argv[optind+1], NULL);
if ((out = sam_open_format(argv[optind+1], wmode, &ga.out)) == NULL) {
print_error_errno("fixmate", "cannot open output file");
goto fail;
}
if (ga.nthreads > 0) {
if (!(p.pool = hts_tpool_init(ga.nthreads))) {
fprintf(stderr, "Error creating thread pool\n");
goto fail;
}
hts_set_opt(in, HTS_OPT_THREAD_POOL, &p);
hts_set_opt(out, HTS_OPT_THREAD_POOL, &p);
}
// run
res = bam_mating_core(in, out, remove_reads, proper_pair_check, add_ct, mate_score);
// cleanup
sam_close(in);
if (sam_close(out) < 0) {
fprintf(stderr, "[bam_mating] error while closing output file\n");
res = 1;
}
if (p.pool) hts_tpool_destroy(p.pool);
sam_global_args_free(&ga);
return res;
fail:
if (in) sam_close(in);
if (out) sam_close(out);
if (p.pool) hts_tpool_destroy(p.pool);
sam_global_args_free(&ga);
return 1;
}
// Set the initial state
static state_t* init(parsed_opts_t* opts)
{
state_t* retval = calloc(sizeof(state_t), 1);
if (!retval) {
fprintf(pysam_stderr, "Out of memory");
return NULL;
}
retval->merged_input_file = sam_open_format(opts->merged_input_name, "rb", &opts->ga.in);
if (!retval->merged_input_file) {
fprintf(pysam_stderr, "Could not open input file (%s)\n", opts->merged_input_name);
free(retval);
return NULL;
}
retval->merged_input_header = sam_hdr_read(retval->merged_input_file);
if (retval->merged_input_header == NULL) {
fprintf(pysam_stderr, "Could not read header for file '%s'\n",
opts->merged_input_name);
cleanup_state(retval, false);
return NULL;
}
if (opts->unaccounted_name) {
if (opts->unaccounted_header_name) {
samFile* hdr_load = sam_open_format(opts->unaccounted_header_name, "r", &opts->ga.in);
if (!hdr_load) {
fprintf(pysam_stderr, "Could not open unaccounted header file (%s)\n", opts->unaccounted_header_name);
cleanup_state(retval, false);
return NULL;
}
retval->unaccounted_header = sam_hdr_read(hdr_load);
if (retval->unaccounted_header == NULL) {
fprintf(pysam_stderr, "Could not read header for file '%s'\n",
opts->unaccounted_header_name);
cleanup_state(retval, false);
return NULL;
}
sam_close(hdr_load);
} else {
retval->unaccounted_header = bam_hdr_dup(retval->merged_input_header);
}
retval->unaccounted_file = sam_open_format(opts->unaccounted_name, "wb", &opts->ga.out);
if (retval->unaccounted_file == NULL) {
fprintf(pysam_stderr, "Could not open unaccounted output file: %s\n", opts->unaccounted_name);
cleanup_state(retval, false);
return NULL;
}
}
// Open output files for RGs
if (!count_RG(retval->merged_input_header, &retval->output_count, &retval->rg_id)) return NULL;
if (opts->verbose) fprintf(pysam_stderr, "@RG's found %zu\n",retval->output_count);
retval->rg_output_file_name = (char **)calloc(retval->output_count, sizeof(char *));
retval->rg_output_file = (samFile**)calloc(retval->output_count, sizeof(samFile*));
retval->rg_output_header = (bam_hdr_t**)calloc(retval->output_count, sizeof(bam_hdr_t*));
retval->rg_hash = kh_init_c2i();
if (!retval->rg_output_file_name || !retval->rg_output_file || !retval->rg_output_header || !retval->rg_hash) {
fprintf(pysam_stderr, "Could not allocate memory for output file array. Out of memory?");
cleanup_state(retval, false);
return NULL;
}
char* dirsep = strrchr(opts->merged_input_name, '/');
char* input_base_name = strdup(dirsep? dirsep+1 : opts->merged_input_name);
if (!input_base_name) {
fprintf(pysam_stderr, "Out of memory\n");
cleanup_state(retval, false);
return NULL;
}
char* extension = strrchr(input_base_name, '.');
if (extension) *extension = '\0';
size_t i;
for (i = 0; i < retval->output_count; i++) {
char* output_filename = NULL;
output_filename = expand_format_string(opts->output_format_string,
input_base_name,
retval->rg_id[i], i,
&opts->ga.out);
if ( output_filename == NULL ) {
fprintf(pysam_stderr, "Error expanding output filename format string.\n");
cleanup_state(retval, false);
free(input_base_name);
return NULL;
}
retval->rg_output_file_name[i] = output_filename;
retval->rg_output_file[i] = sam_open_format(output_filename, "wb", &opts->ga.out);
if (retval->rg_output_file[i] == NULL) {
fprintf(pysam_stderr, "Could not open output file: %s\n", output_filename);
cleanup_state(retval, false);
free(input_base_name);
return NULL;
}
// Record index in hash
int ret;
khiter_t iter = kh_put_c2i(retval->rg_hash, retval->rg_id[i], &ret);
kh_val(retval->rg_hash,iter) = i;
// Set and edit header
retval->rg_output_header[i] = bam_hdr_dup(retval->merged_input_header);
if ( !filter_header_rg(retval->rg_output_header[i], retval->rg_id[i]) ) {
fprintf(pysam_stderr, "Could not rewrite header for file: %s\n", output_filename);
cleanup_state(retval, false);
free(input_base_name);
return NULL;
}
}
free(input_base_name);
return retval;
}
