effectively revert to the viterbi version. The forward realignment gives too many...

[samtools.git] / bam2bcf_indel.c

diff --git a/bam2bcf_indel.c b/bam2bcf_indel.c
index 7f6e288014cf1e0b627f3e7f2150e6dfbef5a03d..2e0b04dfee68f01032665345ac7961ea7ba1a059 100644 (file)
--- a/bam2bcf_indel.c
+++ b/bam2bcf_indel.c
@@ -236,8 +236,9 @@ int bcf_call_gap_prep(int n, int *n_plp, bam_pileup1_t **plp, int pos, bcf_calla
        bca->indelreg = 0;
        for (t = 0; t < n_types; ++t) {
                int l, ir;
-               ka_param2_t ap = ka_param2_qual;
-               ap.band_width = abs(types[t]) + 3;
+               kpa_par_t apf = { 1e-4, 1e-2, 10 };
+               ka_param2_t apv = ka_param2_qual;
+               apf.bw = apv.band_width = abs(types[t]) + 3;
                // compute indelreg
                if (types[t] == 0) ir = 0;
                else if (types[t] > 0) ir = est_indelreg(pos, ref, types[t], &inscns[t*max_ins]);
@@ -264,6 +265,7 @@ int bcf_call_gap_prep(int n, int *n_plp, bam_pileup1_t **plp, int pos, bcf_calla
                                bam_pileup1_t *p = plp[s] + i;
                                int qbeg, qend, tbeg, tend, sc;
                                uint8_t *seq = bam1_seq(p->b);
+                               // FIXME: the following skips soft clips, but using them may be more sensitive.
                                // determine the start and end of sequences for alignment
                                qbeg = tpos2qpos(&p->b->core, bam1_cigar(p->b), left,  0, &tbeg);
                                qend = tpos2qpos(&p->b->core, bam1_cigar(p->b), right, 1, &tend);
@@ -274,16 +276,24 @@ int bcf_call_gap_prep(int n, int *n_plp, bam_pileup1_t **plp, int pos, bcf_calla
                                // write the query sequence
                                for (l = qbeg; l < qend; ++l)
                                        query[l - qbeg] = bam_nt16_nt4_table[bam1_seqi(seq, l)];
-                               // do alignment; this takes most of computing time for indel calling
+                               // do alignment; this is the bottleneck
                                if (0) {
-                                       uint8_t *qq = calloc(qend - qbeg, 1);
-                                       for (l = 0; l < qend - qbeg; ++l) qq[l] = 23;
+                                       const uint8_t *qual = bam1_qual(p->b), *bq;
+                                       uint8_t *qq = 0;
+                                       qq = calloc(qend - qbeg, 1);
+                                       bq = (uint8_t*)bam_aux_get(p->b, "BQ");
+                                       if (bq) ++bq;
+                                       for (l = qbeg; l < qend; ++l) {
+                                               qq[l - qbeg] = bq? qual[l] + (bq[l] - 33) : qual[l];
+                                               if (qq[l - qbeg] > 30) qq[l - qbeg] = 30;
+                                       }
                                        sc = kpa_glocal((uint8_t*)ref2 + tbeg - left, tend - tbeg + abs(types[t]),
-                                                                       (uint8_t*)query, qend - qbeg, qq, &kpa_par_alt, 0, 0);
+                                                                       (uint8_t*)query, qend - qbeg, qq, &apf, 0, 0);
                                        score[K*n_types + t] = sc;
+                                       free(qq);
                                } else {
                                        sc = ka_global_score((uint8_t*)ref2 + tbeg - left, tend - tbeg + abs(types[t]),
-                                                                                (uint8_t*)query, qend - qbeg, &ap);
+                                                                                (uint8_t*)query, qend - qbeg, &apv);
                                        score[K*n_types + t] = -sc;
                                }
 /*