diff --git a/hit.c b/hit.c
index 5a3b3c7..69261a9 100644
--- a/hit.c
+++ b/hit.c
@@ -134,13 +134,13 @@ void mm_sync_regs(void *km, int n_regs, mm_reg1_t *regs) // keep mm_reg1_t::{id,
 	kfree(km, tmp);
 }
 
-void mm_select_sub(void *km, float mask_level, float pri_ratio, int *n_, mm_reg1_t *r)
+void mm_select_sub(void *km, float mask_level, float pri_ratio, int best_n, int *n_, mm_reg1_t *r)
 {
 	if (pri_ratio > 0.0f && *n_ > 0) {
-		int i, k, n = *n_;
+		int i, k, n = *n_, n_2nd = 0;
 		for (i = k = 0; i < n; ++i)
-			if (r[i].parent == i || r[i].score >= r[r[i].parent].score * pri_ratio)
-				r[k++] = r[i];
+			if (r[i].parent == i) r[k++] = r[i];
+			else if (r[i].score >= r[r[i].parent].score * pri_ratio && n_2nd++ < best_n) r[k++] = r[i];
 			else if (r[i].p) free(r[i].p);
 		if (k != n) mm_sync_regs(km, k, r); // removing hits requires sync()
 		*n_ = k;
@@ -192,9 +192,9 @@ int mm_squeeze_a(void *km, int n_regs, mm_reg1_t *regs, mm128_t *a)
 	return as;
 }
 
-void mm_join_long(void *km, const mm_mapopt_t *opt, int qlen, int n_regs, mm_reg1_t *regs, mm128_t *a)
+void mm_join_long(void *km, const mm_mapopt_t *opt, int qlen, int *n_regs_, mm_reg1_t *regs, mm128_t *a)
 {
-	int i, n_aux;
+	int i, n_aux, n_regs = *n_regs_, n_drop = 0;
 	uint64_t *aux;
 
 	if (n_regs < 2) return; // nothing to join
@@ -228,8 +228,21 @@ void mm_join_long(void *km, const mm_mapopt_t *opt, int qlen, int n_regs, mm_reg
 		r0->cnt += r1->cnt, r0->score += r1->score;
 		mm_reg_set_coor(r0, qlen, a);
 		r1->cnt = 0;
+		r1->parent = r0->id;
+		++n_drop;
 	}
 	kfree(km, aux);
+
+	if (n_drop > 0) { // then fix the hits hierarchy
+		for (i = 0; i < n_regs; ++i) { // adjust the mm_reg1_t::parent
+			mm_reg1_t *r = &regs[i];
+			if (r->parent >= 0 && r->id != r->parent) { // fix for secondary hits only
+				if (regs[r->parent].parent >= 0 && regs[r->parent].parent != r->parent)
+					r->parent = regs[r->parent].parent;
+			}
+		}
+		mm_filter_regs(km, opt, n_regs_, regs);
+	}
 }
 
 void mm_set_mapq(int n_regs, mm_reg1_t *regs)
diff --git a/main.c b/main.c
index 6804548..95fcaa8 100644
--- a/main.c
+++ b/main.c
@@ -10,7 +10,7 @@
 #include "minimap.h"
 #include "mmpriv.h"
 
-#define MM_VERSION "2.0-r141-pre"
+#define MM_VERSION "2.0-r142-pre"
 
 void liftrlimit()
 {
@@ -68,7 +68,7 @@ int main(int argc, char *argv[])
 	mm_realtime0 = realtime();
 	mm_mapopt_init(&opt);
 
-	while ((c = getopt_long(argc, argv, "aw:k:t:r:f:Vv:g:I:d:ST:s:x:Hcp:M:n:z:A:B:O:E:m:D:", long_options, &long_idx)) >= 0) {
+	while ((c = getopt_long(argc, argv, "aw:k:t:r:f:Vv:g:I:d:ST:s:x:Hcp:M:n:z:A:B:O:E:m:D:N:", long_options, &long_idx)) >= 0) {
 		if (c == 'w') w = atoi(optarg);
 		else if (c == 'k') k = atoi(optarg);
 		else if (c == 'H') is_hpc = 1;
@@ -78,6 +78,7 @@ int main(int argc, char *argv[])
 		else if (c == 't') n_threads = atoi(optarg);
 		else if (c == 'v') mm_verbose = atoi(optarg);
 		else if (c == 'g') opt.max_gap = atoi(optarg);
+		else if (c == 'N') opt.best_n = atoi(optarg);
 		else if (c == 'p') opt.pri_ratio = atof(optarg);
 		else if (c == 'D') opt.min_seedcov_ratio = atof(optarg);
 		else if (c == 'M') opt.mask_level = atof(optarg);
diff --git a/map.c b/map.c
index 53c85db..bb6a2c9 100644
--- a/map.c
+++ b/map.c
@@ -22,8 +22,9 @@ void mm_mapopt_init(mm_mapopt_t *opt)
 	opt->max_chain_skip = 15;
 	opt->min_seedcov_ratio = 0.0f;
 
-	opt->pri_ratio = 2.0f;
 	opt->mask_level = 0.5f;
+	opt->pri_ratio = 2.0f;
+	opt->best_n = 5;
 
 	opt->max_join_long = 20000;
 	opt->max_join_short = 2000;
@@ -240,14 +241,14 @@ mm_reg1_t *mm_map_frag(const mm_mapopt_t *opt, const mm_idx_t *mi, mm_tbuf_t *b,
 	*n_regs = n_u;
 	if (!(opt->flag & MM_F_AVA)) { // don't choose primary mapping(s) for read overlap
 		mm_set_parent(b->km, opt->mask_level, *n_regs, regs);
-		mm_select_sub(b->km, opt->mask_level, opt->pri_ratio, n_regs, regs);
-		mm_join_long(b->km, opt, qlen, *n_regs, regs, a); // TODO: this can be applied to all-vs-all in principle
+		mm_select_sub(b->km, opt->mask_level, opt->pri_ratio, opt->best_n, n_regs, regs);
+		mm_join_long(b->km, opt, qlen, n_regs, regs, a); // TODO: this can be applied to all-vs-all in principle
 	}
 	if (opt->flag & MM_F_CIGAR) {
 		regs = mm_align_skeleton(b->km, opt, mi, qlen, seq, n_regs, regs, a); // this calls mm_filter_regs()
 		if (!(opt->flag & MM_F_AVA)) {
 			mm_update_parent(b->km, opt->mask_level, *n_regs, regs);
-			mm_select_sub(b->km, opt->mask_level, opt->pri_ratio, n_regs, regs);
+			mm_select_sub(b->km, opt->mask_level, opt->pri_ratio, opt->best_n, n_regs, regs);
 		}
 	} else mm_filter_regs(b->km, opt, n_regs, regs);
 	mm_set_mapq(*n_regs, regs);
diff --git a/minimap.h b/minimap.h
index 51e99e7..af0cd00 100644
--- a/minimap.h
+++ b/minimap.h
@@ -77,8 +77,9 @@ typedef struct {
 	int min_chain_score;
 	float min_seedcov_ratio;
 
-	float pri_ratio;
 	float mask_level;
+	float pri_ratio;
+	int best_n;
 
 	int max_join_long, max_join_short;
 	int min_join_flank_sc;
diff --git a/mmpriv.h b/mmpriv.h
index 62d088a..e706ece 100644
--- a/mmpriv.h
+++ b/mmpriv.h
@@ -41,9 +41,9 @@ void mm_split_reg(mm_reg1_t *r, mm_reg1_t *r2, int n, int qlen, mm128_t *a);
 void mm_sync_regs(void *km, int n_regs, mm_reg1_t *regs);
 void mm_set_parent(void *km, float mask_level, int n, mm_reg1_t *r);
 void mm_update_parent(void *km, float mask_level, int n, mm_reg1_t *r);
-void mm_select_sub(void *km, float mask_level, float pri_ratio, int *n_, mm_reg1_t *r);
+void mm_select_sub(void *km, float mask_level, float pri_ratio, int best_n, int *n_, mm_reg1_t *r);
 void mm_filter_regs(void *km, const mm_mapopt_t *opt, int *n_regs, mm_reg1_t *regs);
-void mm_join_long(void *km, const mm_mapopt_t *opt, int qlen, int n_regs, mm_reg1_t *regs, mm128_t *a);
+void mm_join_long(void *km, const mm_mapopt_t *opt, int qlen, int *n_regs, mm_reg1_t *regs, mm128_t *a);
 void mm_set_mapq(int n_regs, mm_reg1_t *regs);
 
 #ifdef __cplusplus