if (theta <= 0.) theta = CALL_DEFTHETA;
bca = calloc(1, sizeof(bcf_callaux_t));
bca->capQ = 60;
- bca->openQ = 40; bca->extQ = 20; bca->tandemQ = 80;
+ bca->openQ = 40; bca->extQ = 20; bca->tandemQ = 100;
bca->min_baseQ = min_baseQ;
bca->e = errmod_init(1. - theta);
return bca;
{
if (bca == 0) return;
errmod_destroy(bca->e);
- free(bca->bases); free(bca);
+ free(bca->bases); free(bca->inscns); free(bca);
}
/* ref_base is the 4-bit representation of the reference base. It is
* negative if we are looking at an indel. */
memset(r, 0, sizeof(bcf_callret1_t));
for (i = n = 0; i < _n; ++i) {
const bam_pileup1_t *p = pl + i;
- int q, b, mapQ, baseQ, is_diff, min_dist;
+ int q, b, mapQ, baseQ, is_diff, min_dist, seqQ;
// set base
if (p->is_del || p->is_refskip || (p->b->core.flag&BAM_FUNMAP)) continue;
baseQ = q = is_indel? p->aux&0xff : (int)bam1_qual(p->b)[p->qpos]; // base/indel quality
+ seqQ = is_indel? (p->aux>>8&0xff) : 99;
if (q < bca->min_baseQ) continue;
+ if (q > seqQ) q = seqQ;
mapQ = p->b->core.qual < bca->capQ? p->b->core.qual : bca->capQ;
if (q > mapQ) q = mapQ;
if (q > 63) q = 63;
b = bam_nt16_nt4_table[b? b : ref_base]; // b is the 2-bit base
is_diff = (ref4 < 4 && b == ref4)? 0 : 1;
} else {
- b = p->aux>>8&0x3f;
+ b = p->aux>>16&0x3f;
is_diff = (b != 0);
}
bca->bases[n++] = q<<5 | (int)bam1_strand(p->b)<<4 | b;
return 0;
}
-int bcf_call2bcf(int tid, int pos, bcf_call_t *bc, bcf1_t *b, bcf_callret1_t *bcr, int is_SP)
+int bcf_call2bcf(int tid, int pos, bcf_call_t *bc, bcf1_t *b, bcf_callret1_t *bcr, int is_SP,
+ const bcf_callaux_t *bca, const char *ref)
{
extern double kt_fisher_exact(int n11, int n12, int n21, int n22, double *_left, double *_right, double *two);
kstring_t s;
- int i;
+ int i, j;
b->n_smpl = bc->n;
b->tid = tid; b->pos = pos; b->qual = 0;
s.s = b->str; s.m = b->m_str; s.l = 0;
kputc('\0', &s);
- if (bc->ori_ref < 0) {
- kputc('N', &s); kputc('\0', &s);
- kputs("<INDEL>", &s); kputc('\0', &s);
- } else {
+ if (bc->ori_ref < 0) { // an indel
+ // write REF
+ kputc(ref[pos], &s);
+ for (j = 0; j < bca->indelreg; ++j) kputc(ref[pos+1+j], &s);
+ kputc('\0', &s);
+ // write ALT
+ kputc(ref[pos], &s);
+ for (i = 1; i < 4; ++i) {
+ if (bc->a[i] < 0) break;
+ if (i > 1) {
+ kputc(',', &s); kputc(ref[pos], &s);
+ }
+ if (bca->indel_types[bc->a[i]] < 0) { // deletion
+ for (j = -bca->indel_types[bc->a[i]]; j < bca->indelreg; ++j)
+ kputc(ref[pos+1+j], &s);
+ } else { // insertion; cannot be a reference unless a bug
+ char *inscns = &bca->inscns[bc->a[i] * bca->maxins];
+ for (j = 0; j < bca->indel_types[bc->a[i]]; ++j)
+ kputc("ACGTN"[(int)inscns[j]], &s);
+ for (j = 0; j < bca->indelreg; ++j) kputc(ref[pos+1+j], &s);
+ }
+ }
+ kputc('\0', &s);
+ } else { // a SNP
kputc("ACGTN"[bc->ori_ref], &s); kputc('\0', &s);
for (i = 1; i < 5; ++i) {
if (bc->a[i] < 0) break;
}
kputc('\0', &s);
// INFO
+ if (bc->ori_ref < 0) kputs("INDEL;", &s);
kputs("I16=", &s);
for (i = 0; i < 16; ++i) {
if (i) kputc(',', &s);