diff --git a/ksw2_extd2_sse.c b/ksw2_extd2_sse.c
index ca4af4d..c678a11 100644
--- a/ksw2_extd2_sse.c
+++ b/ksw2_extd2_sse.c
@@ -61,7 +61,7 @@ void ksw_extd2_sse(void *km, int qlen, const uint8_t *query, int tlen, const uin
 	int with_cigar = !(flag&KSW_EZ_SCORE_ONLY), approx_max = !!(flag&KSW_EZ_APPROX_MAX);
 	int32_t *H = 0, H0 = 0, last_H0_t = 0;
 	uint8_t *qr, *sf, *mem, *mem2 = 0;
-	__m128i q_, q2_, qe_, qe2_, zero_, sc_mch_, sc_mis_, m1_;
+	__m128i q_, q2_, qe_, qe2_, zero_, sc_mch_, sc_mis_, m1_, sc_N_;
 	__m128i *u, *v, *x, *y, *x2, *y2, *s, *p = 0;
 
 	ksw_reset_extz(ez);
@@ -76,6 +76,7 @@ void ksw_extd2_sse(void *km, int qlen, const uint8_t *query, int tlen, const uin
 	qe2_    = _mm_set1_epi8(q2 + e2);
 	sc_mch_ = _mm_set1_epi8(mat[0]);
 	sc_mis_ = _mm_set1_epi8(mat[1]);
+	sc_N_   = _mm_set1_epi8(-e2);
 	m1_     = _mm_set1_epi8(m - 1); // wildcard
 
 	if (w < 0) w = tlen > qlen? tlen : qlen;
@@ -162,10 +163,11 @@ void ksw_extd2_sse(void *km, int qlen, const uint8_t *query, int tlen, const uin
 				tmp = _mm_cmpeq_epi8(sq, st);
 #ifdef __SSE4_1__
 				tmp = _mm_blendv_epi8(sc_mis_, sc_mch_, tmp);
+				tmp = _mm_blendv_epi8(tmp,     sc_N_,   mask);
 #else
-				tmp = _mm_or_si128(_mm_andnot_si128(tmp, sc_mis_), _mm_and_si128(tmp, sc_mch_));
+				tmp = _mm_or_si128(_mm_andnot_si128(tmp,  sc_mis_), _mm_and_si128(tmp,  sc_mch_));
+				tmp = _mm_or_si128(_mm_andnot_si128(mask, tmp),     _mm_and_si128(mask, sc_N_));
 #endif
-				tmp = _mm_andnot_si128(mask, tmp);
 				_mm_storeu_si128((__m128i*)((int8_t*)s + t), tmp);
 			}
 		} else {
diff --git a/ksw2_exts2_sse.c b/ksw2_exts2_sse.c
index 66149d2..82a0369 100644
--- a/ksw2_exts2_sse.c
+++ b/ksw2_exts2_sse.c
@@ -59,7 +59,7 @@ void ksw_exts2_sse(void *km, int qlen, const uint8_t *query, int tlen, const uin
 	int with_cigar = !(flag&KSW_EZ_SCORE_ONLY), approx_max = !!(flag&KSW_EZ_APPROX_MAX);
 	int32_t *H = 0, H0 = 0, last_H0_t = 0;
 	uint8_t *qr, *sf, *mem, *mem2 = 0;
-	__m128i q_, q2_, qe_, zero_, sc_mch_, sc_mis_, m1_;
+	__m128i q_, q2_, qe_, zero_, sc_mch_, sc_mis_, sc_N_, m1_;
 	__m128i *u, *v, *x, *y, *x2, *s, *p = 0, *donor, *acceptor;
 
 	ksw_reset_extz(ez);
@@ -71,6 +71,7 @@ void ksw_exts2_sse(void *km, int qlen, const uint8_t *query, int tlen, const uin
 	qe_     = _mm_set1_epi8(q + e);
 	sc_mch_ = _mm_set1_epi8(mat[0]);
 	sc_mis_ = _mm_set1_epi8(mat[1]);
+	sc_N_   = _mm_set1_epi8(-e);
 	m1_     = _mm_set1_epi8(m - 1); // wildcard
 
 	tlen_ = (tlen + 15) / 16;
@@ -159,10 +160,11 @@ void ksw_exts2_sse(void *km, int qlen, const uint8_t *query, int tlen, const uin
 				tmp = _mm_cmpeq_epi8(sq, st);
 #ifdef __SSE4_1__
 				tmp = _mm_blendv_epi8(sc_mis_, sc_mch_, tmp);
+				tmp = _mm_blendv_epi8(tmp,     sc_N_,   mask);
 #else
-				tmp = _mm_or_si128(_mm_andnot_si128(tmp, sc_mis_), _mm_and_si128(tmp, sc_mch_));
+				tmp = _mm_or_si128(_mm_andnot_si128(tmp,  sc_mis_), _mm_and_si128(tmp,  sc_mch_));
+				tmp = _mm_or_si128(_mm_andnot_si128(mask, tmp),     _mm_and_si128(mask, sc_N_));
 #endif
-				tmp = _mm_andnot_si128(mask, tmp);
 				_mm_storeu_si128((__m128i*)((int8_t*)s + t), tmp);
 			}
 		} else {
diff --git a/ksw2_extz2_sse.c b/ksw2_extz2_sse.c
index 3889955..e8f71a4 100644
--- a/ksw2_extz2_sse.c
+++ b/ksw2_extz2_sse.c
@@ -50,7 +50,7 @@ void ksw_extz2_sse(void *km, int qlen, const uint8_t *query, int tlen, const uin
 	int with_cigar = !(flag&KSW_EZ_SCORE_ONLY), approx_max = !!(flag&KSW_EZ_APPROX_MAX);
 	int32_t *H = 0, H0 = 0, last_H0_t = 0;
 	uint8_t *qr, *sf, *mem, *mem2 = 0;
-	__m128i q_, qe2_, zero_, flag1_, flag2_, flag8_, flag16_, sc_mch_, sc_mis_, m1_, max_sc_;
+	__m128i q_, qe2_, zero_, flag1_, flag2_, flag8_, flag16_, sc_mch_, sc_mis_, sc_N_, m1_, max_sc_;
 	__m128i *u, *v, *x, *y, *s, *p = 0;
 
 	ksw_reset_extz(ez);
@@ -65,6 +65,7 @@ void ksw_extz2_sse(void *km, int qlen, const uint8_t *query, int tlen, const uin
 	flag16_ = _mm_set1_epi8(0x10);
 	sc_mch_ = _mm_set1_epi8(mat[0]);
 	sc_mis_ = _mm_set1_epi8(mat[1]);
+	sc_N_   = _mm_set1_epi8(-e);
 	m1_     = _mm_set1_epi8(m - 1); // wildcard
 	max_sc_ = _mm_set1_epi8(mat[0] + (q + e) * 2);
 
@@ -130,10 +131,11 @@ void ksw_extz2_sse(void *km, int qlen, const uint8_t *query, int tlen, const uin
 				tmp = _mm_cmpeq_epi8(sq, st);
 #ifdef __SSE4_1__
 				tmp = _mm_blendv_epi8(sc_mis_, sc_mch_, tmp);
+				tmp = _mm_blendv_epi8(tmp,     sc_N_,   mask);
 #else
-				tmp = _mm_or_si128(_mm_andnot_si128(tmp, sc_mis_), _mm_and_si128(tmp, sc_mch_));
+				tmp = _mm_or_si128(_mm_andnot_si128(tmp,  sc_mis_), _mm_and_si128(tmp,  sc_mch_));
+				tmp = _mm_or_si128(_mm_andnot_si128(mask, tmp),     _mm_and_si128(mask, sc_N_));
 #endif
-				tmp = _mm_andnot_si128(mask, tmp);
 				_mm_storeu_si128((__m128i*)((uint8_t*)s + t), tmp);
 			}
 		} else {
diff --git a/main.c b/main.c
index db363fd..35e395a 100644
--- a/main.c
+++ b/main.c
@@ -6,7 +6,7 @@
 #include "mmpriv.h"
 #include "getopt.h"
 
-#define MM_VERSION "2.2-r490-dirty"
+#define MM_VERSION "2.2-r491-dirty"
 
 #ifdef __linux__
 #include <sys/resource.h>
diff --git a/map.c b/map.c
index 32a0685..fef7322 100644
--- a/map.c
+++ b/map.c
@@ -74,10 +74,12 @@ int mm_set_opt(const char *preset, mm_idxopt_t *io, mm_mapopt_t *mo)
 		io->is_hpc = 0, io->k = 19, io->w = 19;
 		mo->a = 1, mo->b = 19, mo->q = 39, mo->q2 = 81, mo->e = 3, mo->e2 = 1, mo->zdrop = 200;
 		mo->min_dp_max = 200;
+		mo->n_best = 50;
 	} else if (strcmp(preset, "asm10") == 0) {
 		io->is_hpc = 0, io->k = 19, io->w = 19;
 		mo->a = 1, mo->b = 9, mo->q = 16, mo->q2 = 41, mo->e = 2, mo->e2 = 1, mo->zdrop = 200;
 		mo->min_dp_max = 200;
+		mo->n_best = 50;
 	} else if (strcmp(preset, "short") == 0 || strcmp(preset, "sr") == 0) {
 		io->is_hpc = 0, io->k = 21, io->w = 11;
 		mo->flag |= MM_F_SR | MM_F_FRAG_MODE | MM_F_NO_PRINT_2ND;