transcript merge implemented (--merge option)

Makefile

@@ -43,7 +43,7 @@ endif
 
 
 BASEFLAGS  := -Wall -Wextra ${SEARCHDIRS} $(MARCH) -D_FILE_OFFSET_BITS=64 \
--D_LARGEFILE_SOURCE -fno-strict-aliasing -fno-exceptions -fno-rtti
+-D_LARGEFILE_SOURCE -fno-strict-aliasing -fno-exceptions -fno-rtti -D_DARWIN_C_SOURCE
 
 # C/C++ linker
 
@@ -103,7 +103,7 @@ endif
 
 
 
-OBJS += rlink.o tablemaker.o
+OBJS += rlink.o tablemaker.o tmerge.o
 
 .PHONY : all debug clean cleanall cleanAll allclean release nothreads
 all:     stringtie
@@ -119,6 +119,7 @@ nothreads: stringtie
 ${GDIR}/GBam.o : $(GDIR)/GBam.h
 stringtie.o : $(GDIR)/GBitVec.h $(GDIR)/GHash.hh $(GDIR)/GBam.h
 rlink.o : rlink.h tablemaker.h $(GDIR)/GBam.h $(GDIR)/GBitVec.h
+tmerge.o : rlink.h tmerge.h
 tablemaker.o : tablemaker.h rlink.h
 ${BAM}/libbam.a: 
 	cd ${BAM} && make lib

README

@@ -43,6 +43,7 @@ The following optional parameters can be specified (use -h/--help to get the
 usage message):
 
  --version : print current version at stdout
+ -h print this usage message
  -G reference annotation to use for guiding the assembly process (GTF/GFF3)
  -l name prefix for output transcripts (default: STRG)
  -f minimum isoform fraction (default: 0.1)
@@ -65,6 +66,24 @@ usage message):
     created under the directory path given as <dir_path>
  -e only estimates the abundance of given reference transcripts (requires -G)
  -x do not assemble any transcripts on the given reference sequence(s)
+Transcript merge usage mode:
+ stringtie --merge [Options] { gtf_list | strg1.gtf ...}
+With this option StringTie will assemble transcripts from multiple\n\
+input files generating a unified non-redundant set of isoforms. In this\n\
+usage mode the following options are available:\n\
+  -G <guide_gff>   reference annotation to include in the merging (GTF/GFF3)\n\
+  -o <out_gtf>     output file name for the merged transcripts GTF\n\
+                    (default: stdout)\n\
+  -m <min_len>     minimum input transcript length to include in the merge\n\
+                    (default: 50)\n\
+  -c <min_cov>     minimum input transcript coverage to include in the merge\n\
+                    (default: 0)\n\
+  -F <min_fpkm>    minimum input transcript FPKM to include in the merge\n\
+                    (default: 0)\n\
+  -T <min_tpm>     minimum input transcript TPM to include in the merge\n\
+                    (default: 0)\n\
+  -f <min_iso>     minimum isoform fraction (default: 0.01)\n\
+  -l <label>       name prefix for output transcripts (default: MSTRG)\n\
 
 Input files
 -----------
@@ -74,9 +93,9 @@ This file contains spliced read alignments such as the ones produced by TopHat o
 A text file in SAM format should be converted to BAM and sorted using the 
 samtools program:
 
-  samtools view -S -b input.sam | samtools sort - input.sorted
+  samtools view -Su alns.sam | samtools sort - alns.sorted
 
-The file resulted from the above command (input.sorted.bam) can be used 
+The file resulted from the above command (alns.sorted.bam) can be used 
 directly as input to StringTie. 
 
 Any SAM spliced read alignment (a read alignment across at least one junction)

gclib/GBase.cpp

@@ -1,6 +1,7 @@
 #include "GBase.h"
 #include <stdarg.h>
 #include <ctype.h>
+#include <errno.h>
 
 #ifndef S_ISDIR
 #define S_ISDIR(mode)  (((mode) & S_IFMT) == S_IFDIR)
@@ -188,7 +189,7 @@ int Gstrcmp(const char* a, const char* b, int n) {
 
 }
 
-int G_mkdir(const char* path, int perms=0775) {
+int G_mkdir(const char* path, int perms=(S_IRWXU | S_IRGRP | S_IXGRP | S_IROTH | S_IXOTH) ) {
  #ifdef __WIN32__
      return _mkdir(path);
  #else
@@ -201,6 +202,18 @@ int G_mkdir(const char* path, int perms=0775) {
 }
 
 
+void Gmktempdir(char* templ) {
+#ifdef __WIN32__
+  int blen=strlen(templ);
+  if (_mktemp_s(templ, blen)!=0)
+	  GError("Error creating temp dir %s!\n", templ);
+#else
+  char* cdir=mkdtemp(templ);
+  if (cdir==NULL)
+	  GError("Error creating temp dir %s!\n", templ);
+#endif
+}
+
 int Gmkdir(const char *path, bool recursive, int perms) {
 	if (path==NULL || path[0]==0) return -1;
 	if (!recursive) return G_mkdir(path, perms);
@@ -222,7 +235,10 @@ int Gmkdir(const char *path, bool recursive, int perms) {
 		*psep=0; //now gpath is the path up to this /
 		ss=psep; ++ss; //ss repositioned just after the /
 		// create current level
-		if (fileExists(gpath)!=1 && G_mkdir(gpath, perms)!=0) {
+		int mkdir_err=0;
+		if (fileExists(gpath)!=1 && (mkdir_err=G_mkdir(gpath, perms))!=0 ) {
+			if (mkdir_err!=0)
+				 GMessage("Warning: failed at mkdir(%s): %s\n",gpath,strerror(errno));
 			GFREE(gpath);
 			return -1;
 		}

gclib/GBase.h

@@ -172,7 +172,7 @@ inline int iround(double x) {
 }
 
 int Gmkdir(const char *path, bool recursive=true, int perms=0775);
-
+void Gmktempdir(char* templ);
 
 /****************************************************************************/