#include <math.h>
#include <stdint.h>
+#include <assert.h>
#include "bam.h"
#include "kstring.h"
#include "bam2bcf.h"
bca->min_frac = 0.002;
bca->min_support = 1;
bca->per_sample_flt = 0;
- return bca;
+ bca->npos = 100;
+ bca->ref_pos = calloc(bca->npos, sizeof(int));
+ bca->alt_pos = calloc(bca->npos, sizeof(int));
+ return bca;
+}
+
+
+static int get_position(const bam_pileup1_t *p, int *len)
+{
+ int icig, n_tot_bases = 0, iread = 0, edist = p->qpos + 1;
+ for (icig=0; icig<p->b->core.n_cigar; icig++)
+ {
+ // Conversion from uint32_t to MIDNSHP
+ // 0123456
+ // MIDNSHP
+ int cig = bam1_cigar(p->b)[icig] & BAM_CIGAR_MASK;
+ int ncig = bam1_cigar(p->b)[icig] >> BAM_CIGAR_SHIFT;
+ if ( cig==0 )
+ {
+ n_tot_bases += ncig;
+ iread += ncig;
+ }
+ else if ( cig==1 )
+ {
+ n_tot_bases += ncig;
+ iread += ncig;
+ }
+ else if ( cig==4 )
+ {
+ iread += ncig;
+ if ( iread<=p->qpos ) edist -= ncig;
+ }
+ }
+ *len = n_tot_bases;
+ return edist;
}
void bcf_call_destroy(bcf_callaux_t *bca)
{
if (bca == 0) return;
errmod_destroy(bca->e);
+ if (bca->npos) { free(bca->ref_pos); free(bca->alt_pos); bca->npos = 0; }
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. */
int bcf_call_glfgen(int _n, const bam_pileup1_t *pl, int ref_base, bcf_callaux_t *bca, bcf_callret1_t *r)
{
- static int *var_pos = NULL, nvar_pos = 0;
int i, n, ref4, is_indel, ori_depth = 0;
memset(r, 0, sizeof(bcf_callret1_t));
if (ref_base >= 0) {
if (min_dist > p->qpos) min_dist = p->qpos;
if (min_dist > CAP_DIST) min_dist = CAP_DIST;
r->anno[1<<2|is_diff<<1|0] += baseQ;
- r->anno[1<<2|is_diff<<1|1] += baseQ * baseQ; // FIXME: signed int is not enough for thousands of samples
+ r->anno[1<<2|is_diff<<1|1] += baseQ * baseQ;
r->anno[2<<2|is_diff<<1|0] += mapQ;
- r->anno[2<<2|is_diff<<1|1] += mapQ * mapQ; // FIXME: signed int is not enough for thousands of samples
+ r->anno[2<<2|is_diff<<1|1] += mapQ * mapQ;
r->anno[3<<2|is_diff<<1|0] += min_dist;
r->anno[3<<2|is_diff<<1|1] += min_dist * min_dist;
+
+ // collect read positions for ReadPosBias
+ int len, pos = get_position(p, &len);
+ int epos = (double)pos/(len+1) * bca->npos;
+ if ( bam1_seqi(bam1_seq(p->b),p->qpos) == ref_base )
+ bca->ref_pos[epos]++;
+ else
+ bca->alt_pos[epos]++;
}
r->depth = n; r->ori_depth = ori_depth;
// glfgen
errmod_cal(bca->e, n, 5, bca->bases, r->p);
+ return r->depth;
+}
- // Calculate the Variant Distance Bias (make it optional?)
- if ( nvar_pos < _n ) {
- nvar_pos = _n;
- var_pos = realloc(var_pos,sizeof(int)*nvar_pos);
- }
- int alt_dp=0, read_len=0;
- for (i=0; i<_n; i++) {
- const bam_pileup1_t *p = pl + i;
- if ( bam1_seqi(bam1_seq(p->b),p->qpos) == ref_base )
- continue;
+double mann_whitney_1947(int n, int m, int U)
+{
+ if (U<0) return 0;
+ if (n==0||m==0) return U==0 ? 1 : 0;
+ return (double)n/(n+m)*mann_whitney_1947(n-1,m,U-m) + (double)m/(n+m)*mann_whitney_1947(n,m-1,U);
+}
- var_pos[alt_dp] = p->qpos;
- if ( (bam1_cigar(p->b)[0]&BAM_CIGAR_MASK)==4 )
- var_pos[alt_dp] -= bam1_cigar(p->b)[0]>>BAM_CIGAR_SHIFT;
+void calc_ReadPosBias(bcf_callaux_t *bca, bcf_call_t *call)
+{
+ int i, nref = 0, nalt = 0;
+ unsigned long int U = 0;
+ for (i=0; i<bca->npos; i++)
+ {
+ nref += bca->ref_pos[i];
+ nalt += bca->alt_pos[i];
+ U += nref*bca->alt_pos[i];
+ bca->ref_pos[i] = 0;
+ bca->alt_pos[i] = 0;
+ }
+#if 0
+//todo
+ double var = 0, avg = (double)(nref+nalt)/bca->npos;
+ for (i=0; i<bca->npos; i++)
+ {
+ double ediff = bca->ref_pos[i] + bca->alt_pos[i] - avg;
+ var += ediff*ediff;
+ bca->ref_pos[i] = 0;
+ bca->alt_pos[i] = 0;
+ }
+ call->read_pos.avg = avg;
+ call->read_pos.var = sqrt(var/bca->npos);
+ call->read_pos.dp = nref+nalt;
+#endif
+ if ( !nref || !nalt )
+ {
+ call->read_pos_bias = -1;
+ return;
+ }
- alt_dp++;
- read_len += p->b->core.l_qseq;
+ if ( nref>=8 || nalt>=8 )
+ {
+ // normal approximation
+ double mean = ((double)nref*nalt+1.0)/2.0;
+ double var2 = (double)nref*nalt*(nref+nalt+1.0)/12.0;
+ double z = (U-mean)/sqrt(var2);
+ call->read_pos_bias = z;
+ //fprintf(stderr,"nref=%d nalt=%d U=%ld mean=%e var=%e zval=%e\n", nref,nalt,U,mean,sqrt(var2),call->read_pos_bias);
}
- float mvd=0;
- int j;
- n=0;
- for (i=0; i<alt_dp; i++) {
- for (j=0; j<i; j++) {
- mvd += abs(var_pos[i] - var_pos[j]);
- n++;
+ else
+ {
+ double p = mann_whitney_1947(nalt,nref,U);
+ // biased form claimed by GATK to behave better empirically
+ // double var2 = (1.0+1.0/(nref+nalt+1.0))*(double)nref*nalt*(nref+nalt+1.0)/12.0;
+ double var2 = (double)nref*nalt*(nref+nalt+1.0)/12.0;
+ double z;
+ if ( p >= 1./sqrt(var2*2*M_PI) ) z = 0; // equal to mean
+ else
+ {
+ if ( U >= nref*nalt/2. ) z = sqrt(-2*log(sqrt(var2*2*M_PI)*p));
+ else z = -sqrt(-2*log(sqrt(var2*2*M_PI)*p));
}
+ call->read_pos_bias = z;
+ //fprintf(stderr,"nref=%d nalt=%d U=%ld p=%e var2=%e zval=%e\n", nref,nalt,U, p,var2,call->read_pos_bias);
}
- r->mvd[0] = n ? mvd/n : 0;
- r->mvd[1] = alt_dp;
- r->mvd[2] = alt_dp ? read_len/alt_dp : 0;
-
- return r->depth;
}
-
-void calc_vdb(int n, const bcf_callret1_t *calls, bcf_call_t *call)
+float mean_diff_to_prob(float mdiff, int dp, int readlen)
{
- // Variant distance bias. Samples merged by means of DP-weighted average.
-
- float weight=0, tot_prob=0;
-
- int i;
- for (i=0; i<n; i++)
+ if ( dp==2 )
{
- int mvd = calls[i].mvd[0];
- int dp = calls[i].mvd[1];
- int read_len = calls[i].mvd[2];
+ if ( mdiff==0 )
+ return (2.0*readlen + 4.0*(readlen-1.0))/((float)readlen*readlen);
+ else
+ return 8.0*(readlen - 4.0*mdiff)/((float)readlen*readlen);
+ }
- if ( dp<2 ) continue;
+ // This is crude empirical approximation and is not very accurate for
+ // shorter read lengths (<100bp). There certainly is a room for
+ // improvement.
+ const float mv[24][2] = { {0,0}, {0,0}, {0,0},
+ { 9.108, 4.934}, { 9.999, 3.991}, {10.273, 3.485}, {10.579, 3.160},
+ {10.828, 2.889}, {11.014, 2.703}, {11.028, 2.546}, {11.244, 2.391},
+ {11.231, 2.320}, {11.323, 2.138}, {11.403, 2.123}, {11.394, 1.994},
+ {11.451, 1.928}, {11.445, 1.862}, {11.516, 1.815}, {11.560, 1.761},
+ {11.544, 1.728}, {11.605, 1.674}, {11.592, 1.652}, {11.674, 1.613},
+ {11.641, 1.570} };
- float prob = 0;
- if ( dp==2 )
- {
- // Exact formula
- prob = (mvd==0) ? 1.0/read_len : (read_len-mvd)*2.0/read_len/read_len;
- }
- else if ( dp==3 )
- {
- // Sin, quite accurate approximation
- float mu = read_len/2.9;
- prob = mvd>2*mu ? 0 : sin(mvd*3.14/2/mu) / (4*mu/3.14);
- }
- else
- {
- // Scaled gaussian curve, crude approximation, but behaves well. Using fixed depth for bigger depths.
- if ( dp>5 )
- dp = 5;
- float sigma2 = (read_len/1.9/(dp+1)) * (read_len/1.9/(dp+1));
- float norm = 1.125*sqrt(2*3.14*sigma2);
- float mu = read_len/2.9;
- if ( mvd < mu )
- prob = exp(-(mvd-mu)*(mvd-mu)/2/sigma2)/norm;
- else
- prob = exp(-(mvd-mu)*(mvd-mu)/3.125/sigma2)/norm;
- }
+ float m, v;
+ if ( dp>=24 )
+ {
+ m = readlen/8.;
+ if (dp>100) dp = 100;
+ v = 1.476/(0.182*pow(dp,0.514));
+ v = v*(readlen/100.);
+ }
+ else
+ {
+ m = mv[dp][0];
+ v = mv[dp][1];
+ m = m*readlen/100.;
+ v = v*readlen/100.;
+ v *= 1.2; // allow more variability
+ }
+ return 1.0/(v*sqrt(2*M_PI)) * exp(-0.5*((mdiff-m)/v)*((mdiff-m)/v));
+}
- //fprintf(stderr,"dp=%d mvd=%d read_len=%d -> prob=%f\n", dp,mvd,read_len,prob);
- tot_prob += prob*dp;
- weight += dp;
+void calc_vdb(bcf_callaux_t *bca, bcf_call_t *call)
+{
+ int i, dp = 0;
+ float mean_pos = 0, mean_diff = 0;
+ for (i=0; i<bca->npos; i++)
+ {
+ if ( !bca->alt_pos[i] ) continue;
+ dp += bca->alt_pos[i];
+ int j = i<bca->npos/2 ? i : bca->npos - i;
+ mean_pos += bca->alt_pos[i]*j;
+ }
+ if ( dp<2 )
+ {
+ call->vdb = -1;
+ return;
+ }
+ mean_pos /= dp;
+ for (i=0; i<bca->npos; i++)
+ {
+ if ( !bca->alt_pos[i] ) continue;
+ int j = i<bca->npos/2 ? i : bca->npos - i;
+ mean_diff += bca->alt_pos[i] * fabs(j - mean_pos);
}
- tot_prob = weight ? tot_prob/weight : 1;
- //fprintf(stderr,"prob=%f\n", tot_prob);
- call->vdb = tot_prob;
+ mean_diff /= dp;
+ call->vdb = mean_diff_to_prob(mean_diff, dp, bca->npos);
}
-int bcf_call_combine(int n, const bcf_callret1_t *calls, int ref_base /*4-bit*/, bcf_call_t *call)
+/**
+ * bcf_call_combine() - sets the PL array and VDB, RPB annotations, finds the top two alleles
+ * @n: number of samples
+ * @calls: each sample's calls
+ * @bca: auxiliary data structure for holding temporary values
+ * @ref_base: the reference base
+ * @call: filled with the annotations
+ */
+int bcf_call_combine(int n, const bcf_callret1_t *calls, bcf_callaux_t *bca, int ref_base /*4-bit*/, bcf_call_t *call)
{
int ref4, i, j, qsum[4];
int64_t tmp;
for (j = 0; j < 16; ++j) call->anno[j] += calls[i].anno[j];
}
- calc_vdb(n, calls, call);
+ calc_vdb(bca, call);
+ calc_ReadPosBias(bca, call);
return 0;
}
if (i) kputc(',', &s);
kputw(bc->anno[i], &s);
}
+ //ksprintf(&s,";RPS=%d,%f,%f", bc->read_pos.dp,bc->read_pos.avg,bc->read_pos.var);
ksprintf(&s,";QS=%f,%f,%f,%f", bc->qsum[0],bc->qsum[1],bc->qsum[2],bc->qsum[3]);
- if (bc->vdb != 1)
- ksprintf(&s, ";VDB=%.4f", bc->vdb);
+ if (bc->vdb != -1)
+ ksprintf(&s, ";VDB=%e", bc->vdb);
+ if (bc->read_pos_bias != -1 )
+ ksprintf(&s, ";RPB=%e", bc->read_pos_bias);
kputc('\0', &s);
// FMT
kputs("PL", &s);
int min_support, max_support; // for collecting indel candidates
double min_frac, max_frac; // for collecting indel candidates
int per_sample_flt; // indel filtering strategy
+ int *ref_pos, *alt_pos, npos; // for ReadPosBias
// for internal uses
int max_bases;
int indel_types[4];
int maxins, indelreg;
+ int read_len;
char *inscns;
uint16_t *bases;
errmod_t *e;
typedef struct {
int depth, n_supp, ori_depth, qsum[4];
- int anno[16];
+ unsigned int anno[16];
float p[25];
- int mvd[3]; // mean variant distance, number of variant reads, average read length
} bcf_callret1_t;
typedef struct {
float qsum[4];
int n, n_alleles, shift, ori_ref, unseen;
int n_supp; // number of supporting non-reference reads
- int anno[16], depth, ori_depth;
+ unsigned int anno[16], depth, ori_depth;
uint8_t *PL;
float vdb; // variant distance bias
+ float read_pos_bias;
+ struct { float avg, var; int dp; } read_pos;
} bcf_call_t;
#ifdef __cplusplus
bcf_callaux_t *bcf_call_init(double theta, int min_baseQ);
void bcf_call_destroy(bcf_callaux_t *bca);
int bcf_call_glfgen(int _n, const bam_pileup1_t *pl, int ref_base, bcf_callaux_t *bca, bcf_callret1_t *r);
- int bcf_call_combine(int n, const bcf_callret1_t *calls, int ref_base /*4-bit*/, bcf_call_t *call);
+ int bcf_call_combine(int n, const bcf_callret1_t *calls, bcf_callaux_t *bca, int ref_base /*4-bit*/, bcf_call_t *call);
int bcf_call2bcf(int tid, int pos, bcf_call_t *bc, bcf1_t *b, bcf_callret1_t *bcr, int fmt_flag,
const bcf_callaux_t *bca, const char *ref);
int bcf_call_gap_prep(int n, int *n_plp, bam_pileup1_t **plp, int pos, bcf_callaux_t *bca, const char *ref,
return max_i - pos;
}
+/*
+ * @n: number of samples
+ */
int bcf_call_gap_prep(int n, int *n_plp, bam_pileup1_t **plp, int pos, bcf_callaux_t *bca, const char *ref,
const void *rghash)
{
* reduces the power because sometimes, substitutions caused by
* indels are not distinguishable from true mutations. Multiple
* sequence realignment helps to increase the power.
+ *
+ * Masks mismatches present in at least 70% of the reads with 'N'.
*/
{ // construct per-sample consensus
int L = right - left + 1, max_i, max2_i;
if ((double)(max2&0xffff) / ((max2&0xffff) + (max2>>16)) >= 0.7) max2_i = -1;
if (max_i >= 0) r[max_i] = 15;
if (max2_i >= 0) r[max2_i] = 15;
-// for (i = 0; i < right - left; ++i) fputc("=ACMGRSVTWYHKDBN"[(int)r[i]], stderr); fputc('\n', stderr);
+ //for (i = 0; i < right - left; ++i) fputc("=ACMGRSVTWYHKDBN"[(int)r[i]], stderr); fputc('\n', stderr);
}
free(ref0); free(cns);
}
}
}
// construct the consensus sequence
- max_ins = types[n_types - 1]; // max_ins is at least 0
+ max_ins = types[n_types - 1]; // max_ins is at least 0
if (max_ins > 0) {
- int *inscns_aux = calloc(4 * n_types * max_ins, sizeof(int));
+ int *inscns_aux = calloc(5 * n_types * max_ins, sizeof(int));
// count the number of occurrences of each base at each position for each type of insertion
for (t = 0; t < n_types; ++t) {
if (types[t] > 0) {
uint8_t *seq = bam1_seq(p->b);
for (k = 1; k <= p->indel; ++k) {
int c = bam_nt16_nt4_table[bam1_seqi(seq, p->qpos + k)];
- if (c < 4) ++inscns_aux[(t*max_ins+(k-1))*4 + c];
+ assert(c<5);
+ ++inscns_aux[(t*max_ins+(k-1))*5 + c];
}
}
}
inscns = calloc(n_types * max_ins, 1);
for (t = 0; t < n_types; ++t) {
for (j = 0; j < types[t]; ++j) {
- int max = 0, max_k = -1, *ia = &inscns_aux[(t*max_ins+j)*4];
- for (k = 0; k < 4; ++k)
+ int max = 0, max_k = -1, *ia = &inscns_aux[(t*max_ins+j)*5];
+ for (k = 0; k < 5; ++k)
if (ia[k] > max)
max = ia[k], max_k = k;
inscns[t*max_ins + j] = max? max_k : 4;
+ if ( max_k==4 ) { types[t] = 0; break; } // discard insertions which contain N's
}
}
free(inscns_aux);
#include "sam.h"
#include "faidx.h"
#include "kstring.h"
+#include "sam_header.h"
static inline int printw(int c, FILE *fp)
{
int rflag_require, rflag_filter;
int openQ, extQ, tandemQ, min_support; // for indels
double min_frac; // for indels
- char *reg, *pl_list;
+ char *reg, *pl_list, *fai_fname;
faidx_t *fai;
void *bed, *rghash;
} mplp_conf_t;
}
data[i]->conf = conf;
h_tmp = bam_header_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);
rghash = bcf_call_add_rg(rghash, h_tmp->text, conf->pl_list);
bh->l_smpl = s.l;
bh->sname = malloc(s.l);
memcpy(bh->sname, s.s, s.l);
- bh->txt = malloc(strlen(BAM_VERSION) + 64);
- bh->l_txt = 1 + sprintf(bh->txt, "##samtoolsVersion=%s\n", BAM_VERSION);
- free(s.s);
+ s.l = 0;
+ ksprintf(&s, "##samtoolsVersion=%s\n", BAM_VERSION);
+ if (conf->fai_fname) ksprintf(&s, "##reference=file://%s\n", conf->fai_fname);
+ h->dict = sam_header_parse2(h->text);
+ int nseq;
+ const char *tags[] = {"SN","LN","UR","M5",NULL};
+ char **tbl = sam_header2tbl_n(h->dict, "SQ", tags, &nseq);
+ for (i=0; i<nseq; i++)
+ {
+ ksprintf(&s, "##contig=<ID=%s", tbl[4*i]);
+ if ( tbl[4*i+1] ) ksprintf(&s, ",length=%s", tbl[4*i+1]);
+ if ( tbl[4*i+2] ) ksprintf(&s, ",URL=%s", tbl[4*i+2]);
+ if ( tbl[4*i+3] ) ksprintf(&s, ",md5=%s", tbl[4*i+3]);
+ kputs(">\n", &s);
+ }
+ if (tbl) free(tbl);
+ bh->txt = s.s;
+ bh->l_txt = 1 + s.l;
bcf_hdr_sync(bh);
bcf_hdr_write(bp, bh);
bca = bcf_call_init(-1., conf->min_baseQ);
ref16 = bam_nt16_table[_ref0];
for (i = 0; i < gplp.n; ++i)
bcf_call_glfgen(gplp.n_plp[i], gplp.plp[i], ref16, bca, bcr + i);
- bcf_call_combine(gplp.n, bcr, ref16, &bc);
+ bcf_call_combine(gplp.n, bcr, bca, ref16, &bc);
bcf_call2bcf(tid, pos, &bc, b, bcr, conf->fmt_flag, 0, 0);
bcf_write(bp, bh, b);
bcf_destroy(b);
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) {
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, -1, &bc) >= 0) {
+ if (bcf_call_combine(gplp.n, bcr, bca, -1, &bc) >= 0) {
b = calloc(1, sizeof(bcf1_t));
bcf_call2bcf(tid, pos, &bc, b, bcr, conf->fmt_flag, bca, ref);
bcf_write(bp, bh, b);
case 'f':
mplp.fai = fai_load(optarg);
if (mplp.fai == 0) return 1;
+ mplp.fai_fname = optarg;
break;
case 'd': mplp.max_depth = atoi(optarg); break;
case 'r': mplp.reg = strdup(optarg); break;
int i,j,nals=0;
for (i=0; i<sizeof(int); i++)
if ( mask&1<<i) nals++;
-
if ( b->n_alleles <= nals ) return;
// update ALT, in principle any of the alleles can be removed
bcf_ginfo_t gt;
// check the presence of the GT FMT
if ((s = strstr(b->fmt, ":GT")) == 0) return 0; // no GT or GT is already the first
- if (s[3] != '\0' && s[3] != ':') return 0; // :GTX in fact
+ assert(s[3] == '\0' || s[3] == ':'); // :GTX in fact
tmp = bcf_str2int("GT", 2);
for (i = 0; i < b->n_gi; ++i)
if (b->gi[i].fmt == tmp) break;
// move GT to the first
for (; i > 0; --i) b->gi[i] = b->gi[i-1];
b->gi[0] = gt;
- memmove(b->fmt + 3, b->fmt, s + 1 - b->fmt);
+ if ( s[3]==0 )
+ memmove(b->fmt + 3, b->fmt, s - b->fmt); // :GT
+ else
+ memmove(b->fmt + 3, b->fmt, s - b->fmt + 1); // :GT:
b->fmt[0] = 'G'; b->fmt[1] = 'T'; b->fmt[2] = ':';
return 0;
}
kputs(samples[i], &s); kputc('\0', &s);
}
}
- if (j < n) fprintf(stderr, "<%s> %d samples in the list but not in BCF.", __func__, n - j);
+ if (j < n)
+ {
+ fprintf(stderr, "<%s> %d samples in the list but not in BCF.", __func__, n - j);
+ exit(1);
+ }
kh_destroy(str2id, hash);
h = calloc(1, sizeof(bcf_hdr_t));
*h = *h0;
*_n = 0;
s.l = s.m = 0; s.s = 0;
fp = gzopen(fn, "r");
- if (fp == 0) return 0; // fail to open file
+ if (fp == 0)
+ {
+ // interpret as sample names, not as a file name
+ const char *t = fn, *p = t;
+ while (*t)
+ {
+ t++;
+ if ( *t==',' || !*t )
+ {
+ sam = realloc(sam, sizeof(void*)*(n+1));
+ sam[n] = (char*) malloc(sizeof(char)*(t-p+2));
+ memcpy(sam[n], p, t-p);
+ sam[n][t-p] = 0;
+ sam[n][t-p+1] = 2; // assume diploid
+ p = t+1;
+ n++;
+ }
+ }
+ *_n = n;
+ return sam; // fail to open file
+ }
ks = ks_init(fp);
while (ks_getuntil(ks, 0, &s, &dret) >= 0) {
int l;
kputs("##INFO=<ID=QCHI2,Number=1,Type=Integer,Description=\"Phred scaled PCHI2.\">\n", &str);
if (!strstr(str.s, "##INFO=<ID=RP,"))
kputs("##INFO=<ID=PR,Number=1,Type=Integer,Description=\"# permutations yielding a smaller PCHI2.\">\n", &str);
+ if (!strstr(str.s, "##INFO=<ID=QBD,"))
+ kputs("##INFO=<ID=QBD,Number=1,Type=Float,Description=\"Quality by Depth: QUAL/#reads\">\n", &str);
+ //if (!strstr(str.s, "##INFO=<ID=RPS,"))
+ // kputs("##INFO=<ID=RPS,Number=3,Type=Float,Description=\"Read Position Stats: depth, average, stddev\">\n", &str);
+ if (!strstr(str.s, "##INFO=<ID=RPB,"))
+ kputs("##INFO=<ID=RPB,Number=1,Type=Float,Description=\"Read Position Bias\">\n", &str);
+ if (!strstr(str.s, "##INFO=<ID=MDV,"))
+ kputs("##INFO=<ID=MDV,Number=1,Type=Integer,Description=\"Maximum number of high-quality nonRef reads in samples\">\n", &str);
if (!strstr(str.s, "##INFO=<ID=VDB,"))
- kputs("##INFO=<ID=VDB,Number=1,Type=Float,Description=\"Variant Distance Bias\">\n", &str);
+ kputs("##INFO=<ID=VDB,Number=1,Type=Float,Description=\"Variant Distance Bias (v2) for filtering splice-site artefacts in RNA-seq data. Note: this version may be broken.\">\n", &str);
if (!strstr(str.s, "##FORMAT=<ID=GT,"))
kputs("##FORMAT=<ID=GT,Number=1,Type=String,Description=\"Genotype\">\n", &str);
if (!strstr(str.s, "##FORMAT=<ID=GQ,"))
if ( !(vc.flag&VC_KEEPALT) && (vc.flag&VC_CALL) && vc.min_ma_lrt>=0 )
{
bcf_p1_set_ploidy(b, p1); // could be improved: do this per site to allow pseudo-autosomal regions
- int gts = call_multiallelic_gt(b,p1,vc.min_ma_lrt);
+ int gts = call_multiallelic_gt(b, p1, vc.min_ma_lrt, vc.flag&VC_VARONLY);
if ( gts<=1 && vc.flag & VC_VARONLY ) continue;
}
else if (vc.flag & VC_CALL) { // call variants
extern double kf_gammap(double s, double z);
int test16(bcf1_t *b, anno16_t *a);
-int call_multiallelic_gt(bcf1_t *b, bcf_p1aux_t *ma, double threshold)
+// Wigginton 2005, PMID: 15789306
+// written by Jan Wigginton
+double calc_hwe(int obs_hom1, int obs_hom2, int obs_hets)
+{
+ if (obs_hom1 + obs_hom2 + obs_hets == 0 ) return 1;
+
+ assert(obs_hom1 >= 0 && obs_hom2 >= 0 && obs_hets >= 0);
+
+ int obs_homc = obs_hom1 < obs_hom2 ? obs_hom2 : obs_hom1;
+ int obs_homr = obs_hom1 < obs_hom2 ? obs_hom1 : obs_hom2;
+
+ int rare_copies = 2 * obs_homr + obs_hets;
+ int genotypes = obs_hets + obs_homc + obs_homr;
+
+ double *het_probs = (double*) calloc(rare_copies+1, sizeof(double));
+
+ /* start at midpoint */
+ int mid = rare_copies * (2 * genotypes - rare_copies) / (2 * genotypes);
+
+ /* check to ensure that midpoint and rare alleles have same parity */
+ if ((rare_copies & 1) ^ (mid & 1)) mid++;
+
+ int curr_hets = mid;
+ int curr_homr = (rare_copies - mid) / 2;
+ int curr_homc = genotypes - curr_hets - curr_homr;
+
+ het_probs[mid] = 1.0;
+ double sum = het_probs[mid];
+ for (curr_hets = mid; curr_hets > 1; curr_hets -= 2)
+ {
+ het_probs[curr_hets - 2] = het_probs[curr_hets] * curr_hets * (curr_hets - 1.0) / (4.0 * (curr_homr + 1.0) * (curr_homc + 1.0));
+ sum += het_probs[curr_hets - 2];
+
+ /* 2 fewer heterozygotes for next iteration -> add one rare, one common homozygote */
+ curr_homr++;
+ curr_homc++;
+ }
+
+ curr_hets = mid;
+ curr_homr = (rare_copies - mid) / 2;
+ curr_homc = genotypes - curr_hets - curr_homr;
+ for (curr_hets = mid; curr_hets <= rare_copies - 2; curr_hets += 2)
+ {
+ het_probs[curr_hets + 2] = het_probs[curr_hets] * 4.0 * curr_homr * curr_homc /((curr_hets + 2.0) * (curr_hets + 1.0));
+ sum += het_probs[curr_hets + 2];
+
+ /* add 2 heterozygotes for next iteration -> subtract one rare, one common homozygote */
+ curr_homr--;
+ curr_homc--;
+ }
+ int i;
+ for (i = 0; i <= rare_copies; i++) het_probs[i] /= sum;
+
+ /* p-value calculation for p_hwe */
+ double p_hwe = 0.0;
+ for (i = 0; i <= rare_copies; i++)
+ {
+ if (het_probs[i] > het_probs[obs_hets])
+ continue;
+ p_hwe += het_probs[i];
+ }
+
+ p_hwe = p_hwe > 1.0 ? 1.0 : p_hwe;
+ free(het_probs);
+ return p_hwe;
+
+}
+
+
+static void _bcf1_set_ref(bcf1_t *b, int idp)
+{
+ kstring_t s;
+ int old_n_gi = b->n_gi;
+ s.m = b->m_str; s.l = b->l_str - 1; s.s = b->str;
+ kputs(":GT", &s); kputc('\0', &s);
+ b->m_str = s.m; b->l_str = s.l; b->str = s.s;
+ bcf_sync(b);
+
+ // Call GTs
+ int isample, an = 0;
+ for (isample = 0; isample < b->n_smpl; isample++)
+ {
+ if ( idp>=0 && ((uint16_t*)b->gi[idp].data)[isample]==0 )
+ ((uint8_t*)b->gi[old_n_gi].data)[isample] = 1<<7;
+ else
+ {
+ ((uint8_t*)b->gi[old_n_gi].data)[isample] = 0;
+ an += b->ploidy ? b->ploidy[isample] : 2;
+ }
+ }
+ bcf_fit_alt(b,1);
+ b->qual = 999;
+
+ // Prepare BCF for output: ref, alt, filter, info, format
+ memset(&s, 0, sizeof(kstring_t)); kputc('\0', &s);
+ kputs(b->ref, &s); kputc('\0', &s);
+ kputs(b->alt, &s); kputc('\0', &s); kputc('\0', &s);
+ {
+ ksprintf(&s, "AN=%d;", an);
+ kputs(b->info, &s);
+ anno16_t a;
+ int has_I16 = test16(b, &a) >= 0? 1 : 0;
+ if (has_I16 )
+ {
+ if ( a.is_tested) ksprintf(&s, ";PV4=%.2g,%.2g,%.2g,%.2g", a.p[0], a.p[1], a.p[2], a.p[3]);
+ ksprintf(&s, ";DP4=%d,%d,%d,%d;MQ=%d", a.d[0], a.d[1], a.d[2], a.d[3], a.mq);
+ }
+ kputc('\0', &s);
+ rm_info(&s, "I16=");
+ rm_info(&s, "QS=");
+ }
+ kputs(b->fmt, &s); kputc('\0', &s);
+ free(b->str);
+ b->m_str = s.m; b->l_str = s.l; b->str = s.s;
+ bcf_sync(b);
+}
+
+int call_multiallelic_gt(bcf1_t *b, bcf_p1aux_t *ma, double threshold, int var_only)
{
int nals = 1;
char *p;
}
if ( b->alt[0] && !*p ) nals++;
- if ( nals==1 ) return 1;
-
if ( nals>4 )
{
if ( *b->ref=='N' ) return 0;
exit(1);
}
- // find PL and DP FORMAT indexes
+ // find PL, DV and DP FORMAT indexes
uint8_t *pl = NULL;
- int npl = 0, idp=-1;
- int i;
+ int i, npl = 0, idp = -1, idv = -1;
for (i = 0; i < b->n_gi; ++i)
{
if (b->gi[i].fmt == bcf_str2int("PL", 2))
pl = (uint8_t*)b->gi[i].data;
npl = b->gi[i].len;
}
- if (b->gi[i].fmt == bcf_str2int("DP", 2)) idp=i;
+ else if (b->gi[i].fmt == bcf_str2int("DP", 2)) idp=i;
+ else if (b->gi[i].fmt == bcf_str2int("DV", 2)) idv=i;
+ }
+ if ( nals==1 )
+ {
+ if ( !var_only ) _bcf1_set_ref(b, idp);
+ return 1;
}
if ( !pl ) return -1;
if ( sscanf(p+3,"%lf,%lf,%lf,%lf",&qsum[0],&qsum[1],&qsum[2],&qsum[3])!=4 ) { fprintf(stderr,"Could not parse %s\n",p); exit(1); }
- // Calculate the most likely combination of alleles
+ // Calculate the most likely combination of alleles, remembering the most and second most likely set
int ia,ib,ic, max_als=0, max_als2=0;
- double ref_lk = 0, max_lk = INT_MIN, max_lk2 = INT_MIN, lk_sum = INT_MIN;
+ double ref_lk = 0, max_lk = INT_MIN, max_lk2 = INT_MIN, lk_sum = INT_MIN, lk_sums[3];
for (ia=0; ia<nals; ia++)
{
double lk_tot = 0;
else if ( max_lk2<lk_tot ) { max_lk2 = lk_tot; max_als2 = 1<<ia; }
lk_sum = lk_tot>lk_sum ? lk_tot + log(1+exp(lk_sum-lk_tot)) : lk_sum + log(1+exp(lk_tot-lk_sum));
}
+ lk_sums[0] = lk_sum;
if ( nals>1 )
{
for (ia=0; ia<nals; ia++)
lk_sum = lk_tot>lk_sum ? lk_tot + log(1+exp(lk_sum-lk_tot)) : lk_sum + log(1+exp(lk_tot-lk_sum));
}
}
+ lk_sums[1] = lk_sum;
}
if ( nals>2 )
{
}
}
}
+ lk_sums[2] = lk_sum;
}
// Should we add another allele, does it increase the likelihood significantly?
for (i=0; i<nals; i++) if ( max_als2&1<<i) n2++;
if ( n2<n1 && kf_gammap(1,2.0*(max_lk-max_lk2))<threshold )
{
- max_lk = max_lk2;
+ // the threshold not exceeded, use the second most likely set with fewer alleles
+ max_lk = max_lk2;
max_als = max_als2;
+ n1 = n2;
}
+ lk_sum = lk_sums[n1-1];
// Get the BCF record ready for GT and GQ
kstring_t s;
// Call GTs
int isample, gts=0, ac[4] = {0,0,0,0};
+ int nRR = 0, nAA = 0, nRA = 0, max_dv = 0;
for (isample = 0; isample < b->n_smpl; isample++)
{
int ploidy = b->ploidy ? b->ploidy[isample] : 2;
double *p = pdg + isample*npdg;
int ia, als = 0;
- double lk = 0, lk_sum=0;
+ double lk = 0, lk_s = 0;
for (ia=0; ia<nals; ia++)
{
if ( !(max_als&1<<ia) ) continue;
int iaa = (ia+1)*(ia+2)/2-1;
double _lk = p[iaa]*qsum[ia]*qsum[ia];
if ( _lk > lk ) { lk = _lk; als = ia<<3 | ia; }
- lk_sum += _lk;
+ lk_s += _lk;
}
if ( ploidy==2 )
{
int iab = iaa - ia + ib;
double _lk = 2*qsum[ia]*qsum[ib]*p[iab];
if ( _lk > lk ) { lk = _lk; als = ib<<3 | ia; }
- lk_sum += _lk;
+ lk_s += _lk;
}
}
}
- lk = -log(1-lk/lk_sum)/0.2302585;
- if ( idp>=0 && ((uint16_t*)b->gi[idp].data)[isample]==0 )
+ lk = -log(1-lk/lk_s)/0.2302585;
+ int dp = 0;
+ if ( idp>=0 && (dp=((uint16_t*)b->gi[idp].data)[isample])==0 )
{
+ // no coverage
((uint8_t*)b->gi[old_n_gi].data)[isample] = 1<<7;
((uint8_t*)b->gi[old_n_gi+1].data)[isample] = 0;
continue;
}
+ if ( lk>99 ) lk = 99;
((uint8_t*)b->gi[old_n_gi].data)[isample] = als;
- ((uint8_t*)b->gi[old_n_gi+1].data)[isample] = lk<100 ? (int)lk : 99;
+ ((uint8_t*)b->gi[old_n_gi+1].data)[isample] = (int)lk;
+
+ // For MDV annotation
+ int dv;
+ if ( als && idv>=0 && (dv=((uint16_t*)b->gi[idv].data)[isample]) )
+ {
+ if ( max_dv < dv ) max_dv = dv;
+ }
+
+ // For HWE annotation; multiple ALT alleles treated as one
+ if ( !als ) nRR++;
+ else if ( !(als>>3&7) || !(als&7) ) nRA++;
+ else nAA++;
gts |= 1<<(als>>3&7) | 1<<(als&7);
ac[ als>>3&7 ]++;
ac[ als&7 ]++;
}
+ free(pdg);
bcf_fit_alt(b,max_als);
+ // The VCF spec is ambiguous about QUAL: is it the probability of anything else
+ // (that is QUAL(non-ref) = P(ref)+P(any non-ref other than ALT)) or is it
+ // QUAL(non-ref)=P(ref) and QUAL(ref)=1-P(ref)? Assuming the latter.
+ b->qual = gts>1 ? -4.343*(ref_lk - lk_sum) : -4.343*log(1-exp(ref_lk - lk_sum));
+ if ( b->qual>999 ) b->qual = 999;
// Prepare BCF for output: ref, alt, filter, info, format
memset(&s, 0, sizeof(kstring_t)); kputc('\0', &s);
{
if ( a.is_tested) ksprintf(&s, ";PV4=%.2g,%.2g,%.2g,%.2g", a.p[0], a.p[1], a.p[2], a.p[3]);
ksprintf(&s, ";DP4=%d,%d,%d,%d;MQ=%d", a.d[0], a.d[1], a.d[2], a.d[3], a.mq);
+ ksprintf(&s, ";QBD=%e", b->qual/(a.d[0] + a.d[1] + a.d[2] + a.d[3]));
+ if ( max_dv ) ksprintf(&s, ";MDV=%d", max_dv);
+ }
+ if ( nAA+nRA )
+ {
+ double hwe = calc_hwe(nAA, nRR, nRA);
+ ksprintf(&s, ";HWE=%e", hwe);
}
kputc('\0', &s);
rm_info(&s, "I16=");
kputs(b->fmt, &s); kputc('\0', &s);
free(b->str);
b->m_str = s.m; b->l_str = s.l; b->str = s.s;
- b->qual = gts>1 ? -4.343*(ref_lk - lk_sum) : -4.343*(max_lk - lk_sum);
- if ( b->qual>999 ) b->qual = 999;
bcf_sync(b);
-
- free(pdg);
return gts;
}
void bcf_p1_destroy(bcf_p1aux_t *ma);
void bcf_p1_set_ploidy(bcf1_t *b, bcf_p1aux_t *ma);
int bcf_p1_cal(const bcf1_t *b, int do_contrast, bcf_p1aux_t *ma, bcf_p1rst_t *rst);
- int call_multiallelic_gt(bcf1_t *b, bcf_p1aux_t *ma, double threshold);
+ int call_multiallelic_gt(bcf1_t *b, bcf_p1aux_t *ma, double threshold, int var_only);
int bcf_p1_call_gt(const bcf_p1aux_t *ma, double f0, int k);
void bcf_p1_dump_afs(bcf_p1aux_t *ma);
int bcf_p1_read_prior(bcf_p1aux_t *ma, const char *fn);
return out_dict;
}
+char **sam_header2tbl_n(const void *dict, const char type[2], const char *tags[], int *n)
+{
+ int nout = 0;
+ char **out = NULL;
+
+ *n = 0;
+ list_t *l = (list_t *)dict;
+ if ( !l ) return NULL;
+
+ int i, ntags = 0;
+ while ( tags[ntags] ) ntags++;
+
+ while (l)
+ {
+ HeaderLine *hline = l->data;
+ if ( hline->type[0]!=type[0] || hline->type[1]!=type[1] )
+ {
+ l = l->next;
+ continue;
+ }
+ out = (char**) realloc(out, sizeof(char*)*(nout+1)*ntags);
+ for (i=0; i<ntags; i++)
+ {
+ HeaderTag *key = header_line_has_tag(hline, tags[i]);
+ if ( !key )
+ {
+ out[nout*ntags+i] = NULL;
+ continue;
+ }
+ out[nout*ntags+i] = key->value;
+ }
+ nout++;
+ l = l->next;
+ }
+ *n = nout;
+ return out;
+}
void *sam_header2key_val(void *iter, const char type[2], const char key_tag[2], const char value_tag[2], const char **key, const char **value);
char **sam_header2list(const void *_dict, char type[2], char key_tag[2], int *_n);
+ /*
+ // Usage example
+ int i, j, n;
+ const char *tags[] = {"SN","LN","UR","M5",NULL};
+ void *dict = sam_header_parse2(bam->header->text);
+ char **tbl = sam_header2tbl_n(h->dict, "SQ", tags, &n);
+ for (i=0; i<n; i++)
+ {
+ for (j=0; j<4; j++)
+ if ( tbl[4*i+j] ) printf("\t%s", tbl[4*i+j]);
+ else printf("-");
+ printf("\n");
+ }
+ if (tbl) free(tbl);
+ */
+ char **sam_header2tbl_n(const void *dict, const char type[2], const char *tags[], int *n);
+
void *sam_header2tbl(const void *dict, char type[2], char key_tag[2], char value_tag[2]);
const char *sam_tbl_get(void *h, const char *key);
int sam_tbl_size(void *h);
#include <stdio.h>
#include <unistd.h>
#include <math.h>
+#include <inttypes.h>
#include "sam_header.h"
#include "sam.h"
#include "faidx.h"
}
view_end:
- if (is_count && ret == 0) {
- printf("%ld\n", count); // compilers on some platforms may complain about printing int64_t with %ld
- }
+ if (is_count && ret == 0)
+ printf("%" PRId64 "\n", count);
+
// close files, free and return
free(fn_list); free(fn_ref); free(fn_out); free(g_library); free(g_rg); free(fn_rg);
if (g_bed) bed_destroy(g_bed);
-.TH samtools 1 "2 September 2011" "samtools-0.1.18" "Bioinformatics tools"
+.TH samtools 1 "15 March 2013" "samtools-0.1.19" "Bioinformatics tools"
.SH NAME
.PP
samtools - Utilities for the Sequence Alignment/Map (SAM) format
.PP
bcftools view in.bcf chr2:100-200 > out.vcf
.PP
-bcftools view -vc in.bcf > out.vcf 2> out.afs
+bcftools view -Nvm0.99 in.bcf > out.vcf 2> out.afs
.SH DESCRIPTION
.PP
.TP
.B tview
-samtools tview <in.sorted.bam> [ref.fasta]
+samtools tview
+.RB [ \-p
+.IR chr:pos ]
+.RB [ \-s
+.IR STR ]
+.RB [ \-d
+.IR display ]
+.RI <in.sorted.bam>
+.RI [ref.fasta]
Text alignment viewer (based on the ncurses library). In the viewer,
press `?' for help and press `g' to check the alignment start from a
region in the format like `chr10:10,000,000' or `=10,000,000' when
viewing the same reference sequence.
+.B Options:
+.RS
+.TP 14
+.BI -d \ display
+Output as (H)tml or (C)urses or (T)ext
+.TP
+.BI -p \ chr:pos
+Go directly to this position
+.TP
+.BI -s \ STR
+Display only reads from this sample or read group
+.RE
+
.TP
.B mpileup
-.B samtools mpileup
-.RB [ \-EBug ]
+samtools mpileup
+.RB [ \-EBugp ]
.RB [ \-C
.IR capQcoef ]
.RB [ \-r
.I INT
leads to more indel calls. [40]
.TP
+.BI -p
+Apply -m and -F thresholds per sample to increase sensitivity of calling.
+By default both options are applied to reads pooled from all samples.
+.TP
.BI -P \ STR
Comma dilimited list of platforms (determined by
.BR @RG-PL )
.IR mutRate ]
.RB [ \-p
.IR varThres ]
+.RB [ \-m
+.IR varThres ]
.RB [ \-P
.IR prior ]
.RB [ \-1
.BI -i \ FLOAT
Ratio of INDEL-to-SNP mutation rate [0.15]
.TP
+.BI -m \ FLOAT
+New model for improved multiallelic and rare-variant calling. Another
+ALT allele is accepted if P(chi^2) of LRT exceeds the FLOAT threshold. The
+parameter seems robust and the actual value usually does not affect the results
+much; a good value to use is 0.99. This is the recommended calling method. [0]
+.TP
.BI -p \ FLOAT
A site is considered to be a variant if P(ref|D)<FLOAT [0.5]
.TP
CLR int Phred log ratio of genotype likelihoods with and without the trio/pair constraint
UGT string Most probable genotype configuration without the trio constraint
CGT string Most probable configuration with the trio constraint
+VDB float Tests variant positions within reads. Intended for filtering RNA-seq artifacts around splice sites
+RPB float Mann-Whitney rank-sum test for tail distance bias
+HWE float Hardy-Weinberg equilibrium test, Wigginton et al., PMID: 15789306
.TE
.SH EXAMPLES
Collecting indel candidates from reads sequenced by an indel-prone
technology may affect the performance of indel calling.
+Note that there is a new calling model which can be invoked by
+
+ bcftools view -m0.99 ...
+
+which fixes some severe limitations of the default method.
+
+For filtering, best results seem to be achieved by first applying the
+.IR SnpGap
+filter and then applying some machine learning approach
+
+ vcf-annotate -f SnpGap=n
+ vcf filter ...
+
+Both can be found in the
+.B vcftools
+and
+.B htslib
+package (links below).
+
.IP o 2
Derive the allele frequency spectrum (AFS) on a list of sites from multiple individuals:
.SH SEE ALSO
.PP
Samtools website: <http://samtools.sourceforge.net>
+.br
+Samtools latest source: <https://github.com/samtools/samtools>
+.br
+VCFtools website with stable link to VCF specification: <http://vcftools.sourceforge.net>
+.br
+HTSlib website: <https://github.com/samtools/htslib>
projects / samtools.git / commitdiff