diff --git a/CommandLines.cpp b/CommandLines.cpp
index fda2142..c582d87 100644
--- a/CommandLines.cpp
+++ b/CommandLines.cpp
@@ -60,6 +60,7 @@ static ko_longopt_t long_options[] = {
     { "dbg-ovec",     ko_no_argument, 345},
     { "path-max",     ko_required_argument, 346},
     { "path-min",     ko_required_argument, 347},
+    { "trio-dual",     ko_no_argument, 348},
     // { "path-round",     ko_required_argument, 348},
 	{ 0, 0, 0 }
 };
@@ -129,6 +130,8 @@ void Print_H(hifiasm_opt_t* asm_opt)
     fprintf(stderr, "    --t-occ      INT\n");
     fprintf(stderr, "                 forcedly remove unitigs with >INT unexpected haplotype-specific reads;\n");
     fprintf(stderr, "                 ignore graph topology; [%d]\n", asm_opt->trio_flag_occ_thres);
+    fprintf(stderr, "    --trio-dual  utilize homology information to correct trio phasing errors\n");
+
 
     fprintf(stderr, "  Purge-dups:\n");
     fprintf(stderr, "    -l INT       purge level. 0: no purging; 1: light; 2/3: aggressive [0 for trio; 3 for unzip]\n");
@@ -286,6 +289,7 @@ void init_opt(hifiasm_opt_t* asm_opt)
     asm_opt->max_path_drop_rate = 0.6;
     asm_opt->hifi_pst_join = 1;
     asm_opt->ul_pst_join = 1;
+    asm_opt->trio_cov_het_ovlp = -1;
 }
 
 void destory_enzyme(enzyme* f)
@@ -830,6 +834,7 @@ int CommandLine_process(int argc, char *argv[], hifiasm_opt_t* asm_opt)
         else if (c == 345) asm_opt->dbg_ovec_cal = 1;
         else if (c == 346) asm_opt->max_path_drop_rate = atof(opt.arg);
         else if (c == 347) asm_opt->min_path_drop_rate = atof(opt.arg);
+        else if (c == 348) asm_opt->trio_cov_het_ovlp = 1;
         else if (c == 'l') {   ///0: disable purge_dup; 1: purge containment; 2: purge overlap
             asm_opt->purge_level_primary = asm_opt->purge_level_trio = atoi(opt.arg);
         }
diff --git a/CommandLines.h b/CommandLines.h
index 51e5a6b..6cfa400 100644
--- a/CommandLines.h
+++ b/CommandLines.h
@@ -5,7 +5,7 @@
 #include <pthread.h>
 #include <stdint.h>
 
-#define HA_VERSION "0.19.5-r590"
+#define HA_VERSION "0.19.5-r592"
 
 #define VERBOSE 0
 
@@ -131,7 +131,7 @@ typedef struct {
     float dp_e;
     int64_t hg_size;
     float kpt_rate;
-    int64_t infor_cov, s_hap_cov;
+    int64_t infor_cov, s_hap_cov, trio_cov_het_ovlp;
     double ul_error_rate, ul_error_rate_low, ul_error_rate_hpc;
     int32_t ul_ec_round;
     uint8_t is_dbg_het_cnt;
diff --git a/Overlaps.cpp b/Overlaps.cpp
index a8cdb92..a43877c 100644
--- a/Overlaps.cpp
+++ b/Overlaps.cpp
@@ -17034,6 +17034,62 @@ long long gap_fuzz, ug_opt_t *opt)
     0.05, 0.9, max_hang, min_ovlp, gap_fuzz, b_mask_t, NULL, NULL);
 }
 
+void output_bp_trio_graph(asg_t *sg, ma_sub_t* coverage_cut, char* output_file_name, 
+ma_hit_t_alloc* sources, ma_hit_t_alloc* reverse_sources, long long tipsLen, float tip_drop_ratio, long long stops_threshold, 
+R_to_U* ruIndex, float chimeric_rate, float drop_ratio, int max_hang, int min_ovlp, bub_label_t* b_mask_t,
+long long gap_fuzz, ug_opt_t *opt)
+{
+    hic_clean(sg);
+
+    kvec_asg_arc_t_warp new_rtg_edges;
+    kv_init(new_rtg_edges.a); 
+    ma_ug_t *ug = NULL;
+    ug = ma_ug_gen_primary(sg, PRIMARY_LABLE);
+    
+    hap_cov_t *cov = NULL;
+    trans_chain* t_ch = NULL;
+    if((asm_opt.flag & HA_F_VERBOSE_GFA)) t_ch = load_hc_trans(output_file_name);
+    
+    if(!t_ch) {
+        new_rtg_edges.a.n = 0;
+        asg_t *copy_sg = copy_read_graph(sg);
+        ma_ug_t *copy_ug = copy_untig_graph(ug);    
+        ///asm_opt.purge_overlap_len = asm_opt.purge_overlap_len_hic;
+        ///asm_opt.purge_simi_thres = asm_opt.purge_simi_rate_hic;
+        adjust_utg_by_primary(&copy_ug, copy_sg, TRIO_THRES, sources, reverse_sources, coverage_cut, 
+        tipsLen, tip_drop_ratio, stops_threshold, ruIndex, chimeric_rate, drop_ratio, 
+        max_hang, min_ovlp, &new_rtg_edges, &cov, b_mask_t, 1, 0);
+        
+        ma_ug_destroy(copy_ug);
+        asg_destroy(copy_sg);
+
+        if((asm_opt.flag & HA_F_VERBOSE_GFA)) write_trans_chain(cov->t_ch, output_file_name);
+    }
+
+    // clean_u_trans_t_idx_filter_adv(&(cov?cov->t_ch->k_trans:t_ch->k_trans), ug, sg);
+    trio_phasing_refine(ug, sg, cov?&(cov->t_ch->k_trans):&(t_ch->k_trans), opt);
+    
+    if(cov) destory_hap_cov_t(&cov);
+    if(t_ch) destory_trans_chain(&t_ch);
+
+
+    // char* gfa_name = (char*)malloc(strlen(output_file_name)+50);
+    // sprintf(gfa_name, "%s.after.clean_d_utg.noseq.gfa", output_file_name);
+    // FILE* output_file = fopen(gfa_name, "w");
+    // ma_ug_print_simple(ug, sg, coverage_cut, sources, ruIndex, "utg", output_file);
+    // fclose(output_file);
+    // free(gfa_name);
+
+
+
+    ma_ug_destroy(ug);
+    kv_destroy(new_rtg_edges.a); 
+
+
+    output_trio_graph_joint(sg, coverage_cut, output_file_name, sources, reverse_sources, (asm_opt.max_short_tip*2), 0.15, 3, ruIndex, 
+    0.05, 0.9, max_hang, min_ovlp, gap_fuzz, b_mask_t, NULL, NULL);
+}
+
 ma_ug_t* merge_utg(ma_ug_t **dest, ma_ug_t **src)
 {
     asg_t *g_d = (*dest)->g, *g_s = (*src)->g;
@@ -36393,8 +36449,13 @@ ma_sub_t **coverage_cut_ptr, int debug_g)
         if(asm_opt.flag & HA_F_PARTITION) asm_opt.flag -= HA_F_PARTITION;
         // output_trio_graph(sg, coverage_cut, o_file, sources, reverse_sources, (asm_opt.max_short_tip*2), 
         // 0.15, 3, ruIndex, 0.05, 0.9, max_hang_length, mini_overlap_length, 0, gap_fuzz, &uopt, &b_mask_t);
-        output_trio_graph_joint(sg, coverage_cut, o_file, sources, reverse_sources, (asm_opt.max_short_tip*2), 
-        0.15, 3, ruIndex, 0.05, 0.9, max_hang_length, mini_overlap_length, gap_fuzz, &b_mask_t, NULL, NULL);
+        if(asm_opt.trio_cov_het_ovlp > 0) {
+            output_bp_trio_graph(sg, coverage_cut, o_file, sources, reverse_sources, (asm_opt.max_short_tip*2), 
+            0.15, 3, ruIndex, 0.05, 0.9, max_hang_length, mini_overlap_length, &b_mask_t, gap_fuzz, &uopt);
+        } else {
+            output_trio_graph_joint(sg, coverage_cut, o_file, sources, reverse_sources, (asm_opt.max_short_tip*2), 
+            0.15, 3, ruIndex, 0.05, 0.9, max_hang_length, mini_overlap_length, gap_fuzz, &b_mask_t, NULL, NULL);
+        }
     }
     else if(ha_opt_hic(&asm_opt))
     {
diff --git a/gfa_ut.h b/gfa_ut.h
index c39e0d8..8a395df 100644
--- a/gfa_ut.h
+++ b/gfa_ut.h
@@ -40,5 +40,6 @@ void filter_sg_by_ug(asg_t *rg, ma_ug_t *ug, ug_opt_t *uopt);
 void ug_ext_gfa(ug_opt_t *uopt, asg_t *sg, uint32_t max_len);
 void update_sg_uo(asg_t *g, ma_hit_t_alloc *src);
 uint32_t get_arcs(asg_t *g, uint32_t v, uint32_t* idx, uint32_t idx_n);
+uint64_t ug_occ_w(uint64_t is, uint64_t ie, ma_utg_t *u);
 
 #endif
diff --git a/hic.cpp b/hic.cpp
index ac666cd..28d95db 100644
--- a/hic.cpp
+++ b/hic.cpp
@@ -17722,6 +17722,190 @@ void hic_analysis(ma_ug_t *ug, asg_t* read_g, trans_chain* t_ch, ug_opt_t *opt,
     destory_hc_pt_index(ug_index);free(ug_index);
 }
 
+uint64_t ug_occ_hap_w(uint64_t is, uint64_t ie, ma_utg_t *u)
+{
+    uint64_t l, i, us, ue, occ;
+    for (i = l = occ = 0; i < u->n; i++) {
+        us = l; ue = l + Get_READ_LENGTH(R_INF, (u->a[i]>>33));
+        if(is <= us && ie >= ue) {
+            if((R_INF.trio_flag[u->a[i]>>33] == FATHER) || (R_INF.trio_flag[u->a[i]>>33] == MOTHER)) {
+                occ++;
+            }
+        }
+        if(us >= ie) break;
+        l += (uint32_t)u->a[i];
+    }
+    return occ;
+}
+
+void trio_phasing_refine(ma_ug_t *iug, asg_t* sg, kv_u_trans_t *ref, ug_opt_t *opt)
+{
+    ma_ug_t *ug = copy_untig_graph(iug); uint8_t *bf = NULL; 
+    uint64_t ug_n0 = ug->g->n_seq, k, i, ul, cis_n, trans_n, w_n, tot_hap, tot_r, frid, mrid, flag; 
+    kv_u_trans_t in; memset(&in, 0, sizeof(in)); ma_utg_t *p; u_trans_t *z; int64_t fh, mh;
+    bubble_type *bub = gen_bubble_chain(sg, ug, opt, &bf, 0); free(bf);
+    clean_u_trans_t_idx_filter_adv(ref, ug, sg);
+    
+    ///update ug itself
+    ul = FATHER; frid = ug->g->n_seq; asg_seq_set(ug->g, frid, ul, 0); 
+    kv_pushp(ma_utg_t, ug->u, &p); memset(p, 0, sizeof((*p)));
+    p->s = 0; p->start = UINT32_MAX; p->end = UINT32_MAX; p->len = ul, p->n = p->m = 1; p->circ = 1;
+    kv_roundup32(p->m); p->a = (uint64_t*)malloc(8 * p->m); 
+    p->a[0] = UINT32_MAX; p->a[0] <<= 32; p->a[0] |= ul;
+    
+    ul = MOTHER; mrid = ug->g->n_seq; asg_seq_set(ug->g, mrid, ul, 0); 
+    kv_pushp(ma_utg_t, ug->u, &p); memset(p, 0, sizeof((*p)));
+    p->s = 0; p->start = UINT32_MAX; p->end = UINT32_MAX; p->len = ul, p->n = p->m = 1; p->circ = 1;
+    kv_roundup32(p->m); p->a = (uint64_t*)malloc(8 * p->m); 
+    p->a[0] = UINT32_MAX; p->a[0] <<= 32; p->a[0] |= ul;
+
+    free(ug->g->idx); ug->g->idx = 0; ug->g->is_srt = 0; asg_cleanup(ug->g);
+
+    ///update bubble
+    REALLOC(bub->index, ug->g->n_seq);
+    for (k = ug_n0; k < ug->g->n_seq; k++) bub->index[k] = bub->f_bub+1;
+    
+    bub->b_s_idx.n = ug->g->n_seq; kv_resize(uint64_t, bub->b_s_idx, ug->g->n_seq);
+    for (k = ug_n0; k < bub->b_s_idx.m; k++) bub->b_s_idx.a[k] = (uint64_t)-1;
+
+    ///check if a node has haplotype markers
+    CALLOC(bf, ug->g->n_seq);
+    for (k = cis_n = tot_hap = tot_r = 0; k < ug_n0; k++) {
+        if(ug->g->seq[k].del) continue;
+        p = &(ug->u.a[k]); tot_r += p->n;
+        if(p->n == 0 || p->m == 0) continue;
+        for (i = fh = mh = 0; i < p->n; i++) {
+            if(R_INF.trio_flag[p->a[i]>>33] == FATHER) {
+                tot_hap++; fh = 1;
+            } else if(R_INF.trio_flag[p->a[i]>>33] == MOTHER) {
+                tot_hap++; mh = 1;
+            }
+            if((R_INF.trio_flag[p->a[i]>>33] == FATHER) || (R_INF.trio_flag[p->a[i]>>33] == MOTHER)) {
+                tot_hap++; bf[k] = 1; 
+            }
+        }
+        if(fh > 0) {
+            bf[k] = 1; cis_n+=2;
+        }
+        if(mh > 0) {
+            bf[k] = 1; cis_n+=2;
+        }
+    }
+    for (; k < ug->g->n_seq; k++) bf[k] = 1;///for father/mother nodes
+    
+
+    ///update trans overlaps
+    kv_pushp(u_trans_t, *ref, &z); memset(z, 0, sizeof((*z)));
+    z->f = RC_2; z->nw = (DBL_MAX/2); z->rev = 0;
+    z->qn = ug_n0; z->tn = ug_n0+1; z->del = 0;
+    z->qs = 0; z->qe = ug->g->seq[z->qn].len;
+    z->ts = 0; z->te = ug->g->seq[z->tn].len;
+
+    kv_pushp(u_trans_t, *ref, &z); memset(z, 0, sizeof((*z)));
+    z->f = RC_2; z->nw = (DBL_MAX/2); z->rev = 0;
+    z->qn = ug_n0+1; z->tn = ug_n0; z->del = 0;
+    z->qs = 0; z->qe = ug->g->seq[z->qn].len;
+    z->ts = 0; z->te = ug->g->seq[z->tn].len;
+
+    kt_u_trans_t_idx(ref, ug->g->n_seq);
+    // clean_u_trans_t_idx_adv(ref, ug, sg);
+    // clean_u_trans_t_idx_filter_adv(ref, ug, sg);
+
+    ///gen all links
+    for (k = trans_n = 0; k < ref->n; k++) {
+        if(ref->a[k].del) continue;
+        if((IF_HOM(ref->a[k].qn, (*bub))) || (IF_HOM(ref->a[k].tn, (*bub)))) continue;
+        ///disable bf
+        // if((!bf[ref->a[k].qn]) || (!bf[ref->a[k].tn])) continue;
+        trans_n++;
+    }
+    kv_resize(u_trans_t, in, (trans_n+cis_n)); 
+
+    for (k = in.n = 0; k < ref->n; k++) {
+        if(ref->a[k].del) continue;
+        if((IF_HOM(ref->a[k].qn, (*bub))) || (IF_HOM(ref->a[k].tn, (*bub)))) continue;
+        ///disable bf
+        // if((!bf[ref->a[k].qn]) || (!bf[ref->a[k].tn])) continue;
+
+        kv_pushp(u_trans_t, in, &z); memset(z, 0, sizeof((*z))); w_n = 0;
+        if((ref->a[k].qn < ug_n0) && (ref->a[k].tn < ug_n0)) {
+            w_n += ug_occ_hap_w(ref->a[k].qs, ref->a[k].qe, &(ug->u.a[ref->a[k].qn])) + 
+            ug_occ_hap_w(ref->a[k].ts, ref->a[k].te, &(ug->u.a[ref->a[k].tn]));
+            w_n += ug_occ_w(ref->a[k].qs, ref->a[k].qe, &(ug->u.a[ref->a[k].qn])) + 
+            ug_occ_w(ref->a[k].ts, ref->a[k].te, &(ug->u.a[ref->a[k].tn]));
+        } else {
+            w_n += tot_hap + tot_r;
+        }
+        w_n >>= 1;///wn/4
+        (*z) = ref->a[k]; z->nw = ((w_n)?(w_n):(1));
+    }
+
+    for (k = 0; k < ug_n0; k++) {
+        if(!bf[k]) continue;
+        p = &(ug->u.a[k]);
+        if(p->n == 0 || p->m == 0) continue;
+        fh = mh = 0;
+        for (i = 0; i < p->n; i++) {
+            if(R_INF.trio_flag[p->a[i]>>33] == FATHER) fh--;
+            if(R_INF.trio_flag[p->a[i]>>33] == MOTHER) mh--;
+        }
+        if(fh != 0) {
+            w_n = frid;
+
+            kv_pushp(u_trans_t, in, &z); memset(z, 0, sizeof((*z)));
+            z->f = RC_2; z->nw = fh; z->rev = 0;
+            z->qn = w_n; z->tn = k; z->del = 0;
+            z->qs = 0; z->qe = MIN(ug->g->seq[z->qn].len, ug->g->seq[z->tn].len);
+            z->ts = 0; z->te = MIN(ug->g->seq[z->qn].len, ug->g->seq[z->tn].len);
+
+            kv_pushp(u_trans_t, in, &z); memset(z, 0, sizeof((*z)));
+            z->f = RC_2; z->nw = fh; z->rev = 0;
+            z->qn = k; z->tn = w_n; z->del = 0;
+            z->qs = 0; z->qe = MIN(ug->g->seq[z->qn].len, ug->g->seq[z->tn].len);
+            z->ts = 0; z->te = MIN(ug->g->seq[z->qn].len, ug->g->seq[z->tn].len);
+        }
+
+        if(mh != 0) {
+            w_n = mrid;
+
+            kv_pushp(u_trans_t, in, &z); memset(z, 0, sizeof((*z)));
+            z->f = RC_2; z->nw = mh; z->rev = 0;
+            z->qn = w_n; z->tn = k; z->del = 0;
+            z->qs = 0; z->qe = MIN(ug->g->seq[z->qn].len, ug->g->seq[z->tn].len);
+            z->ts = 0; z->te = MIN(ug->g->seq[z->qn].len, ug->g->seq[z->tn].len);
+
+            kv_pushp(u_trans_t, in, &z); memset(z, 0, sizeof((*z)));
+            z->f = RC_2; z->nw = mh; z->rev = 0;
+            z->qn = k; z->tn = w_n; z->del = 0;
+            z->qs = 0; z->qe = MIN(ug->g->seq[z->qn].len, ug->g->seq[z->tn].len);
+            z->ts = 0; z->te = MIN(ug->g->seq[z->qn].len, ug->g->seq[z->tn].len);
+        }
+    }
+    free(bf);
+    assert(in.n == (trans_n+cis_n));
+    kt_u_trans_t_idx(&in, ug->g->n_seq);
+    // clean_u_trans_t_idx_adv(&in, ug, sg);
+
+    ps_t *s = init_ps_t(11, ug->g->n_seq); s->s[frid] = 1; s->s[mrid] = -1; 
+    mc_solve(NULL, NULL, &in, ug, sg, 0.8, NULL, 0, s->s, 1, bub, ref, 0, 0);
+    // fprintf(stderr, "[M::%s::] s[frid]::%d, s->s[mrid]::%d\n", __func__, s->s[frid], s->s[mrid]);
+    if((s->s[frid] != 0) && (s->s[mrid] != 0) && (s->s[frid] != s->s[mrid])) {
+        memset(R_INF.trio_flag, AMBIGU, R_INF.total_reads * sizeof(uint8_t));
+        for (i = 0; i < ug_n0; i++) {
+            if(ug->g->seq[i].del) continue;
+            flag = AMBIGU;
+            if(s->s[i] == 0) continue;
+            flag = (s->s[i] == s->s[frid]? FATHER:MOTHER);
+            p = &ug->u.a[i];
+            if(p->m == 0) continue;
+            for (k = 0; k < p->n; k++) R_INF.trio_flag[p->a[k]>>33] = flag;
+        }
+    }
+
+    destory_bubbles(bub); free(bub); ma_ug_destroy(ug);
+    destory_ps_t(&s); kv_destroy(in); kv_destroy(in.idx);
+}
+
 spg_t *hic_pre_analysis(ma_ug_t *ug, asg_t* read_g, trans_chain* t_ch, ug_opt_t *opt, kvec_pe_hit **rhits)
 {
     ug_index = NULL;
diff --git a/hic.h b/hic.h
index 61760de..3e66d6e 100644
--- a/hic.h
+++ b/hic.h
@@ -117,5 +117,6 @@ void hic_analysis(ma_ug_t *ug, asg_t* read_g, trans_chain* t_ch, ug_opt_t *opt,
 spg_t *hic_pre_analysis(ma_ug_t *ug, asg_t* read_g, trans_chain* t_ch, ug_opt_t *opt, kvec_pe_hit **rhits);
 void prt_bubble_gfa_adv(FILE *fp, bubble_type *bub, const char* utg_pre, const char* bub_pre, const char* chain_pre);
 void bp_solve(ug_opt_t *opt, kv_u_trans_t *ref, ma_ug_t *ug, asg_t *sg, bubble_type *bub, double cis_rate);
+void trio_phasing_refine(ma_ug_t *ug, asg_t* sg, kv_u_trans_t *ta, ug_opt_t *opt);
 
 #endif
diff --git a/horder.cpp b/horder.cpp
index a691e09..5663f64 100644
--- a/horder.cpp
+++ b/horder.cpp
@@ -3969,8 +3969,8 @@ asg_t *i_rg, ma_ug_t* i_ug, bubble_type* bub, kv_u_trans_t *ref, ug_opt_t *opt,
 
 int cmp_mc_edge_t_w(const void * a, const void * b)
 {
-    if((*(osg_arc_t*)a).nw == (*(osg_arc_t*)b).nw) return 0;
-    return (*(osg_arc_t*)a).nw < (*(osg_arc_t*)b).nw ? -1 : 1;
+    if((*(mc_edge_t*)a).w == (*(mc_edge_t*)b).w) return 0;
+    return (*(mc_edge_t*)a).w < (*(mc_edge_t*)b).w ? -1 : 1;
 }
 
 void cal_chain_arch(scg_t *sg, const mc_match_t *ma, uint32_t v, uint32_t *va, uint32_t vn, uint64_t *idx, asg64_v *srt)
@@ -4303,6 +4303,7 @@ double min_cut, double max_cut, uint64_t cut_round)
         kv_pushp(mc_edge_t, sm, &sp);
         sp->w = sg->g->arc[k].nw; sp->x = k;
     }
+    // fprintf(stderr, "sm.n::%lu\n", (uint64_t)sm.n);
     qsort(sm.a, sm.n, sizeof(mc_edge_t), cmp_mc_edge_t_w);