Merge branch 'dev' into XB

NEWS.md

@@ -1,3 +1,14 @@
+Release 0.7.17 (23 October 2017)
+--------------------------------
+
+This release adds option -q to preserve the mapping quality of split alignment
+with a lower alignment score than the primary alignment. Option -5
+automatically applies -q as well.
+
+(0.7.17: 23 October 2017, r1188)
+
+
+
 Release 0.7.16 (30 July 2017)
 -----------------------------
 

README.md

@@ -1,4 +1,13 @@
 [![Build Status](https://travis-ci.org/lh3/bwa.svg?branch=dev)](https://travis-ci.org/lh3/bwa)
+[![SourceForge Downloads](https://img.shields.io/sourceforge/dt/bio-bwa.svg)](https://sourceforge.net/projects/bio-bwa/files/?source=navbar)
+[![GitHub Downloads](https://img.shields.io/github/downloads/lh3/bwa/total.svg?style=flat)](https://github.com/lh3/bwa/releases)
+
+**Note: [minimap2][minimap2] has replaced BWA-MEM for __PacBio and Nanopore__ read
+alignment.** It retains all major BWA-MEM features, but is ~50 times as fast,
+more versatile, more accurate and produces better base-level alignment.
+
+[minimap2]: https://github.com/lh3/minimap2
+
 ## Getting started
 
 	git clone https://github.com/lh3/bwa.git

bwa.1

@@ -1,4 +1,4 @@
-.TH bwa 1 "30 July 2017" "bwa-0.7.16-r1180" "Bioinformatics tools"
+.TH bwa 1 "23 October 2017" "bwa-0.7.17-r1188" "Bioinformatics tools"
 .SH NAME
 .PP
 bwa - Burrows-Wheeler Alignment Tool
@@ -303,9 +303,14 @@ For compatibility with other BWA commands, this option may also be given as
 .IR FILE .
 [standard ouptut]
 .TP
+.B -q
+ Don't reduce the mapping quality of split alignment of lower alignment score.
+.TP
 .B -5
 For split alignment, mark the segment with the smallest coordinate as the
-primary. This option may help some Hi-C pipelines. By default, BWA-MEM marks
+primary. It automatically applies option
+.B -q
+as well. This option may help some Hi-C pipelines. By default, BWA-MEM marks
 highest scoring segment as primary.
 .TP
 .B -K \ INT

bwamem.c

@@ -1029,7 +1029,8 @@ void mem_reg2sam(const mem_opt_t *opt, const bntseq_t *bns, const uint8_t *pac,
 		if (p->secondary >= 0) q->sub = -1; // don't output sub-optimal score
 		if (l && p->secondary < 0) // if supplementary
 			q->flag |= (opt->flag&MEM_F_NO_MULTI)? 0x10000 : 0x800;
-		if (l && !p->is_alt && q->mapq > aa.a[0].mapq) q->mapq = aa.a[0].mapq;
+		if (!(opt->flag & MEM_F_KEEP_SUPP_MAPQ) && l && !p->is_alt && q->mapq > aa.a[0].mapq)
+			q->mapq = aa.a[0].mapq; // lower mapq for supplementary mappings, unless -5 or -q is applied
 		++l;
 	}
 	if (aa.n == 0) { // no alignments good enough; then write an unaligned record

bwamem.h

@@ -20,7 +20,8 @@ typedef struct __smem_i smem_i;
 #define MEM_F_SOFTCLIP  0x200
 #define MEM_F_SMARTPE   0x400
 #define MEM_F_PRIMARY5  0x800
-#define MEM_F_XB        0x1000
+#define MEM_F_KEEP_SUPP_MAPQ 0x1000
+#define MEM_F_XB        0x2000
 
 typedef struct {
 	int a, b;               // match score and mismatch penalty

fastmap.c

@@ -130,7 +130,7 @@ int main_mem(int argc, char *argv[])
 
 	aux.opt = opt = mem_opt_init();
 	memset(&opt0, 0, sizeof(mem_opt_t));
-	while ((c = getopt(argc, argv, "51paMCSPVYjuk:c:v:s:r:t:R:A:B:O:E:U:w:L:d:T:Q:D:m:I:N:o:f:W:x:G:h:y:K:X:H:")) >= 0) {
+	while ((c = getopt(argc, argv, "51qpaMCSPVYjuk:c:v:s:r:t:R:A:B:O:E:U:w:L:d:T:Q:D:m:I:N:o:f:W:x:G:h:y:K:X:H:")) >= 0) {
 		if (c == 'k') opt->min_seed_len = atoi(optarg), opt0.min_seed_len = 1;
 		else if (c == '1') no_mt_io = 1;
 		else if (c == 'x') mode = optarg;
@@ -147,7 +147,8 @@ int main_mem(int argc, char *argv[])
 		else if (c == 'S') opt->flag |= MEM_F_NO_RESCUE;
 		else if (c == 'Y') opt->flag |= MEM_F_SOFTCLIP;
 		else if (c == 'V') opt->flag |= MEM_F_REF_HDR;
-		else if (c == '5') opt->flag |= MEM_F_PRIMARY5;
+		else if (c == '5') opt->flag |= MEM_F_PRIMARY5 | MEM_F_KEEP_SUPP_MAPQ; // always apply MEM_F_KEEP_SUPP_MAPQ with -5
+		else if (c == 'q') opt->flag |= MEM_F_KEEP_SUPP_MAPQ;
 		else if (c == 'u') opt->flag |= MEM_F_XB;
 		else if (c == 'c') opt->max_occ = atoi(optarg), opt0.max_occ = 1;
 		else if (c == 'd') opt->zdrop = atoi(optarg), opt0.zdrop = 1;
@@ -269,7 +270,8 @@ int main_mem(int argc, char *argv[])
 		fprintf(stderr, "       -H STR/FILE   insert STR to header if it starts with @; or insert lines in FILE [null]\n");
 		fprintf(stderr, "       -o FILE       sam file to output results to [stdout]\n");
 		fprintf(stderr, "       -j            treat ALT contigs as part of the primary assembly (i.e. ignore <idxbase>.alt file)\n");
-		fprintf(stderr, "       -5            for split alignment, take the alignment with the smallest coordiate as primary\n");
+		fprintf(stderr, "       -5            for split alignment, take the alignment with the smallest coordinate as primary\n");
+		fprintf(stderr, "       -q            don't modify mapQ of supplementary alignments\n");
 		fprintf(stderr, "       -K INT        process INT input bases in each batch regardless of nThreads (for reproducibility) []\n");
 		fprintf(stderr, "\n");
 		fprintf(stderr, "       -v INT        verbosity level: 1=error, 2=warning, 3=message, 4+=debugging [%d]\n", bwa_verbose);

main.c

@@ -4,7 +4,7 @@
 #include "utils.h"
 
 #ifndef PACKAGE_VERSION
-#define PACKAGE_VERSION "0.7.16a-r1181"
+#define PACKAGE_VERSION "0.7.17-r1194-dirty"
 #endif
 
 int bwa_fa2pac(int argc, char *argv[]);