X-Git-Url: https://git.donarmstrong.com/?a=blobdiff_plain;f=bam_plcmd.c;h=1d5ea5a82da2076217deb9bf224a5ed12b0655ef;hb=b0ed19dc2d8fb84998d41d6c3588611185707137;hp=79d0e63b8e75a8c6d55c19f6433d9d4c42d159cc;hpb=42e756837f7acb91bfd1134ef16d423e71c913d7;p=samtools.git diff --git a/bam_plcmd.c b/bam_plcmd.c index 79d0e63..1d5ea5a 100644 --- a/bam_plcmd.c +++ b/bam_plcmd.c @@ -1,7 +1,7 @@ #include #include #include -#include "bam.h" +#include "sam.h" #include "faidx.h" #include "bam_maqcns.h" #include "khash.h" @@ -26,7 +26,7 @@ typedef struct { int tid, len, last_pos; int mask; char *ref; - glfFile fp; // for glf output only + glfFile fp_glf; // for glf output only } pu_data_t; char **__bam_get_lines(const char *fn, int *_n); @@ -103,9 +103,9 @@ static int glt3_func(uint32_t tid, uint32_t pos, int n, const bam_pileup1_t *pu, if (d->fai && (int)tid != d->tid) { if (d->ref) { // then write the end mark g3->rtype = GLF3_RTYPE_END; - glf3_write1(d->fp, g3); + glf3_write1(d->fp_glf, g3); } - glf3_ref_write(d->fp, d->h->target_name[tid], d->h->target_len[tid]); // write reference + glf3_ref_write(d->fp_glf, d->h->target_name[tid], d->h->target_len[tid]); // write reference free(d->ref); d->ref = fai_fetch(d->fai, d->h->target_name[tid], &d->len); d->tid = tid; @@ -117,7 +117,7 @@ static int glt3_func(uint32_t tid, uint32_t pos, int n, const bam_pileup1_t *pu, g3->rtype = GLF3_RTYPE_SUB; g3->offset = pos - d->last_pos; d->last_pos = pos; - glf3_write1(d->fp, g3); + glf3_write1(d->fp_glf, g3); if (proposed_indels) r = bam_maqindel(n, pos, d->ido, pu, d->ref, proposed_indels[0], proposed_indels+1); else r = bam_maqindel(n, pos, d->ido, pu, d->ref, 0, 0); @@ -139,7 +139,7 @@ static int glt3_func(uint32_t tid, uint32_t pos, int n, const bam_pileup1_t *pu, g3->max_len = (abs(r->indel1) > abs(r->indel2)? abs(r->indel1) : abs(r->indel2)) + 1; g3->indel_seq[0] = strdup(r->s[0]+1); g3->indel_seq[1] = strdup(r->s[1]+1); - glf3_write1(d->fp, g3); + glf3_write1(d->fp_glf, g3); bam_maqindel_ret_destroy(r); } free(g); @@ -151,7 +151,7 @@ static int pileup_func(uint32_t tid, uint32_t pos, int n, const bam_pileup1_t *p { pu_data_t *d = (pu_data_t*)data; bam_maqindel_ret_t *r = 0; - int i, j, rb, max_mapq = 0, *proposed_indels = 0; + int i, j, rb, rms_mapq = -1, max_mapq = 0, *proposed_indels = 0; uint32_t x; if (d->format & BAM_PLF_GLF) return glt3_func(tid, pos, n, pu, data); @@ -181,6 +181,7 @@ static int pileup_func(uint32_t tid, uint32_t pos, int n, const bam_pileup1_t *p if (ref_q > 255) ref_q = 255; } printf("%c\t%d\t%d\t%d\t", bam_nt16_rev_table[x>>28], x>>8&0xff, ref_q, x>>16&0xff); + rms_mapq = x>>16&0xff; } if ((d->format & (BAM_PLF_CNS|BAM_PLF_INDEL_ONLY)) && d->ref) { if (proposed_indels) @@ -195,7 +196,7 @@ static int pileup_func(uint32_t tid, uint32_t pos, int n, const bam_pileup1_t *p if (p->is_head) printf("^%c", p->b->core.qual > 93? 126 : p->b->core.qual + 33); if (!p->is_del) { int c = bam_nt16_rev_table[bam1_seqi(bam1_seq(p->b), p->qpos)]; - if (toupper(c) == toupper(rb)) c = bam1_strand(p->b)? ',' : '.'; + if (c == '=' || toupper(c) == toupper(rb)) c = bam1_strand(p->b)? ',' : '.'; else c = bam1_strand(p->b)? tolower(c) : toupper(c); putchar(c); if (p->indel > 0) { @@ -235,10 +236,10 @@ static int pileup_func(uint32_t tid, uint32_t pos, int n, const bam_pileup1_t *p if (r->gt < 2) printf("%s/%s\t", r->s[r->gt], r->s[r->gt]); else printf("%s/%s\t", r->s[0], r->s[1]); printf("%d\t%d\t", r->q_cns, r->q_ref); - printf("%d\t%d\t", max_mapq, n); + printf("%d\t%d\t", rms_mapq >= 0? rms_mapq : max_mapq, n); printf("%s\t%s\t", r->s[0], r->s[1]); //printf("%d\t%d\t", r->gl[0], r->gl[1]); - printf("%d\t%d\t%d\t%d\n", r->cnt1, r->cnt2, r->cnt_ambi, r->cnt_anti); + printf("%d\t%d\t%d\n", r->cnt1, r->cnt2, r->cnt_anti); bam_maqindel_ret_destroy(r); } return 0; @@ -260,10 +261,10 @@ int bam_pileup(int argc, char *argv[]) case 'f': fn_fa = strdup(optarg); break; case 'T': d->c->theta = atof(optarg); break; case 'N': d->c->n_hap = atoi(optarg); break; - case 'r': d->c->het_rate = atoi(optarg); break; + case 'r': d->c->het_rate = atof(optarg); break; case 'c': d->format |= BAM_PLF_CNS; break; case 'i': d->format |= BAM_PLF_INDEL_ONLY; break; - case 'm': d->mask = atoi(optarg); break; + case 'm': d->mask = strtol(optarg, 0, 0); break; case 'g': d->format |= BAM_PLF_GLF; break; case 'I': d->ido->q_indel = atoi(optarg); break; case 'G': d->ido->r_indel = atof(optarg); break; @@ -272,7 +273,7 @@ int bam_pileup(int argc, char *argv[]) } if (optind == argc) { fprintf(stderr, "\n"); - fprintf(stderr, "Usage: bamtk pileup [options] |\n\n"); + fprintf(stderr, "Usage: samtools pileup [options] |\n\n"); fprintf(stderr, "Option: -s simple (yet incomplete) pileup format\n"); fprintf(stderr, " -i only show lines/consensus with indels\n"); fprintf(stderr, " -m INT filtering reads with bits in INT [%d]\n", d->mask); @@ -280,7 +281,7 @@ int bam_pileup(int argc, char *argv[]) fprintf(stderr, " -l FILE list of sites at which pileup is output\n"); fprintf(stderr, " -f FILE reference sequence in the FASTA format\n\n"); fprintf(stderr, " -c output the maq consensus sequence\n"); - fprintf(stderr, " -g output in the extended GLT format (suppressing -c/-i/-s)\n"); + fprintf(stderr, " -g output in the GLFv3 format (suppressing -c/-i/-s)\n"); fprintf(stderr, " -T FLOAT theta in maq consensus calling model (for -c/-g) [%f]\n", d->c->theta); fprintf(stderr, " -N INT number of haplotypes in the sample (for -c/-g) [%d]\n", d->c->n_hap); fprintf(stderr, " -r FLOAT prior of a difference between two haplotypes (for -c/-g) [%f]\n", d->c->het_rate); @@ -291,50 +292,38 @@ int bam_pileup(int argc, char *argv[]) return 1; } if (fn_fa) d->fai = fai_load(fn_fa); - free(fn_fa); - bam_maqcns_prepare(d->c); - if (d->format & BAM_PLF_GLF) { + if (d->format & (BAM_PLF_CNS|BAM_PLF_GLF)) bam_maqcns_prepare(d->c); // consensus calling + if (d->format & BAM_PLF_GLF) { // for glf output glf3_header_t *h; h = glf3_header_init(); - d->fp = bgzf_fdopen(fileno(stdout), "w"); - glf3_header_write(d->fp, h); + d->fp_glf = bgzf_fdopen(fileno(stdout), "w"); + glf3_header_write(d->fp_glf, h); glf3_header_destroy(h); } - if (d->fai == 0) + if (d->fai == 0 && (d->format & (BAM_PLF_CNS|BAM_PLF_INDEL_ONLY))) fprintf(stderr, "[bam_pileup] indels will not be called when -f is absent.\n"); - if (fn_list) { // the input is SAM - tamFile fp; - bam1_t *b; - int ret; - bam_plbuf_t *buf = bam_plbuf_init(pileup_func, d); - bam_plbuf_set_mask(buf, d->mask); - d->h = sam_header_read2(fn_list); - if (fn_pos) d->hash = load_pos(fn_pos, d->h); - fp = sam_open(argv[optind]); - b = (bam1_t*)calloc(1, sizeof(bam1_t)); - while ((ret = sam_read1(fp, d->h, b)) >= 0) - bam_plbuf_push(b, buf); - bam_plbuf_push(0, buf); - bam_plbuf_destroy(buf); - bam_destroy1(b); - sam_close(fp); - } else { // the input is BAM - bamFile fp; - fp = (strcmp(argv[optind], "-") == 0)? bam_dopen(fileno(stdin), "r") : bam_open(argv[optind], "r"); - d->h = bam_header_read(fp); + { + samfile_t *fp; + fp = fn_list? samopen(argv[optind], "r", fn_list) : samopen(argv[optind], "rb", 0); + if (fp == 0 || fp->header == 0) { + fprintf(stderr, "[bam_pileup] fail to read the header: non-exisiting file or wrong format.\n"); + return 1; + } + d->h = fp->header; if (fn_pos) d->hash = load_pos(fn_pos, d->h); - bam_pileup_file(fp, d->mask, pileup_func, d); - bam_close(fp); + sampileup(fp, d->mask, pileup_func, d); + samclose(fp); // d->h will be destroyed here } - if (d->format & BAM_PLF_GLF) bgzf_close(d->fp); + + // free + if (d->format & BAM_PLF_GLF) bgzf_close(d->fp_glf); if (fn_pos) { // free the hash table khint_t k; for (k = kh_begin(d->hash); k < kh_end(d->hash); ++k) if (kh_exist(d->hash, k)) free(kh_val(d->hash, k)); kh_destroy(64, d->hash); } - free(fn_pos); free(fn_list); - bam_header_destroy(d->h); + free(fn_pos); free(fn_list); free(fn_fa); if (d->fai) fai_destroy(d->fai); bam_maqcns_destroy(d->c); free(d->ido); free(d->ref); free(d);