bio-velvet_underground 0.0.1 → 0.1.0

data/ext/src/src/run2.c

@@ -68,6 +68,8 @@ static void printUsage()
 	puts("\t-paired_exp_fraction <double>\t: remove all the paired end connections which less than the specified fraction of the expected count (default: 0.1)");
 	puts("\t-shortMatePaired* <yes|no>\t: for mate-pair libraries, indicate that the library might be contaminated with paired-end reads (default no)");
 	puts("\t-conserveLong <yes|no>\t\t: preserve sequences with long reads in them (default no)");
+	puts("\t-clip_tips <yes|no>\t\t: do tip clipping on pre-graph (default yes)");
+	puts("\t-tour_bus <yes|no>\t\t: apply the tour bus algorithm (default yes)");
 	puts("");
 	puts("Output:");
 	puts("\tdirectory/contigs.fa\t\t: fasta file of contigs longer than twice hash length");
@@ -76,7 +78,7 @@ static void printUsage()
 	puts("\tdirectory/velvet_asm.afg\t: (if requested) AMOS compatible assembly file");
 }
 
-int main(int argc, char **argv)
+int velvetg(int argc, char **argv)
 {
 	ReadSet *sequences = NULL;
 	RoadMapArray *rdmaps;
@@ -86,7 +88,9 @@ int main(int argc, char **argv)
 	    *preGraphFilename, *seqFilename, *roadmapFilename,
 	    *lowCovContigsFilename, *highCovContigsFilename;
 	double coverageCutoff = -1;
-	double longCoverageCutoff = -1;
+  boolean doClipTips = true;
+  boolean doTourBus = true;
+  double longCoverageCutoff = -1;
 	double maxCoverageCutoff = -1;
 	double expectedCoverage = -1;
 	Coordinate minContigLength = -1;
@@ -164,7 +168,7 @@ int main(int argc, char **argv)
 		return 0;
 	}
 
-	// Memory allocation 
+	// Memory allocation
 	directory = argv[1];
 	graphFilename = mallocOrExit(strlen(directory) + 100, char);
 	connectedGraphFilename = mallocOrExit(strlen(directory) + 100, char);
@@ -181,9 +185,9 @@ int main(int argc, char **argv)
 		if (arg_index >= argc) {
 			velvetLog("Unusual number of arguments!\n");
 			printUsage();
-#ifdef DEBUG 
+#ifdef DEBUG
 			abort();
-#endif 
+#endif
 			exit(1);
 		}
 
@@ -210,9 +214,9 @@ int main(int argc, char **argv)
 			if (insertLength[0] < 0) {
 				velvetLog("Invalid insert length: %lli\n",
 				       (long long) insertLength[0]);
-#ifdef DEBUG 
+#ifdef DEBUG
 				abort();
-#endif 
+#endif
 				exit(1);
 			}
 		} else if (strcmp(arg, "-ins_length_sd") == 0) {
@@ -221,9 +225,9 @@ int main(int argc, char **argv)
 			if (std_dev[0] < 0) {
 				velvetLog("Invalid std deviation: %lli\n",
 				       (long long) std_dev[0]);
-#ifdef DEBUG 
+#ifdef DEBUG
 				abort();
-#endif 
+#endif
 				exit(1);
 			}
 		} else if (strcmp(arg, "-ins_length_long") == 0) {
@@ -238,9 +242,9 @@ int main(int argc, char **argv)
 			cat = (Category) short_var;
 			if (cat < 1 || cat > CATEGORIES) {
 				velvetLog("Unknown option: %s\n", arg);
-#ifdef DEBUG 
+#ifdef DEBUG
 				abort();
-#endif 
+#endif
 				exit(1);
 			}
 			sscanf(argv[arg_index], "%lli", &longlong_var);
@@ -248,9 +252,9 @@ int main(int argc, char **argv)
 			if (insertLength[cat - 1] < 0) {
 				velvetLog("Invalid insert length: %lli\n",
 				       (long long) insertLength[cat - 1]);
-#ifdef DEBUG 
+#ifdef DEBUG
 				abort();
-#endif 
+#endif
 				exit(1);
 			}
 		} else if (strncmp(arg, "-ins_length", 11) == 0) {
@@ -258,9 +262,9 @@ int main(int argc, char **argv)
 			cat = (Category) short_var;
 			if (cat < 1 || cat > CATEGORIES) {
 				velvetLog("Unknown option: %s\n", arg);
-#ifdef DEBUG 
+#ifdef DEBUG
 				abort();
-#endif 
+#endif
 				exit(1);
 			}
 			sscanf(argv[arg_index], "%lli", &longlong_var);
@@ -268,9 +272,9 @@ int main(int argc, char **argv)
 			if (std_dev[cat - 1] < 0) {
 				velvetLog("Invalid std deviation: %lli\n",
 				       (long long) std_dev[cat - 1]);
-#ifdef DEBUG 
+#ifdef DEBUG
 				abort();
-#endif 
+#endif
 				exit(1);
 			}
 		} else if (strcmp(arg, "-read_trkg") == 0) {
@@ -353,6 +357,12 @@ int main(int argc, char **argv)
 				exit(1);
 			}
 			shadows[cat - 1] = (strcmp(argv[arg_index], "yes") == 0);
+    } else if (strcmp(arg,"-clip_tips") == 0){
+      if (strcmp(argv[arg_index], "no") == 0)
+        doClipTips = false;
+    } else if (strcmp(arg,"-tour_bus") == 0){
+      if (strcmp(argv[arg_index], "no") == 0)
+        doTourBus = false;
 		} else if (strcmp(arg, "--help") == 0) {
 			printUsage();
 			return 0;
@@ -452,7 +462,8 @@ int main(int argc, char **argv)
 
 		sequenceLengths =
 		    getSequenceLengths(sequences, getWordLength(graph));
-		correctGraph(graph, sequenceLengths, sequences->categories, conserveLong);
+    if (doTourBus)
+        correctGraph(graph, sequenceLengths, sequences->categories, conserveLong);
 		exportGraph(graphFilename, graph, sequences->tSequences);
 	} else if ((file = fopen(preGraphFilename, "r")) != NULL) {
 		fclose(file);
@@ -468,7 +479,8 @@ int main(int argc, char **argv)
 				   roadmapFilename, readTracking, accelerationBits);
 		sequenceLengths =
 		    getSequenceLengths(sequences, getWordLength(graph));
-		correctGraph(graph, sequenceLengths, sequences->categories, conserveLong);
+    if (doTourBus)
+        correctGraph(graph, sequenceLengths, sequences->categories, conserveLong);
 		exportGraph(graphFilename, graph, sequences->tSequences);
 	} else if ((file = fopen(roadmapFilename, "r")) != NULL) {
 		fclose(file);
@@ -509,7 +521,7 @@ int main(int argc, char **argv)
 		}
 		preGraph = newPreGraph_pg(rdmaps, seqReadInfo);
 		concatenatePreGraph_pg(preGraph);
-		if (!conserveLong)
+		if (!conserveLong && doClipTips)
 		    clipTips_pg(preGraph);
 		exportPreGraph_pg(preGraphFilename, preGraph);
 		destroyPreGraph_pg(preGraph);
@@ -523,13 +535,14 @@ int main(int argc, char **argv)
 				   roadmapFilename, readTracking, accelerationBits);
 		sequenceLengths =
 		    getSequenceLengths(sequences, getWordLength(graph));
-		correctGraph(graph, sequenceLengths, sequences->categories, conserveLong);
+    if (doTourBus)
+        correctGraph(graph, sequenceLengths, sequences->categories, conserveLong);
 		exportGraph(graphFilename, graph, sequences->tSequences);
 	} else {
 		velvetLog("No Roadmap file to build upon! Please run velveth (see manual)\n");
-#ifdef DEBUG 
+#ifdef DEBUG
 		abort();
-#endif 
+#endif
 		exit(1);
 	}
 
@@ -553,11 +566,11 @@ int main(int argc, char **argv)
 			coverageCutoff = expectedCoverage / 2;
 			estimateCutoff = true;
 		}
-	} else { 
+	} else {
 		estimateCoverage = false;
-		if (coverageCutoff < 0 && estimateCutoff) 
+		if (coverageCutoff < 0 && estimateCutoff)
 			coverageCutoff = estimated_cov(graph, directory) / 2;
-		else 
+		else
 			estimateCutoff = false;
 	}
 
@@ -580,7 +593,7 @@ int main(int argc, char **argv)
 	if (minContigLength < 2 * getWordLength(graph))
 		minContigKmerLength = getWordLength(graph);
 	else
-		minContigKmerLength = minContigLength - getWordLength(graph) + 1;		
+		minContigKmerLength = minContigLength - getWordLength(graph) + 1;
 
 	dubious =
 	    removeLowCoverageNodesAndDenounceDubiousReads(graph,
@@ -599,7 +612,8 @@ int main(int argc, char **argv)
 							  lowCovContigsFilename);
 
 	removeHighCoverageNodes(graph, maxCoverageCutoff, exportFilteredNodes, minContigKmerLength, highCovContigsFilename);
-	clipTipsHard(graph, conserveLong);
+	if (doClipTips)
+		clipTipsHard(graph, conserveLong);
 
 	if (sequences->readCount > 0 && sequences->categories[0] == REFERENCE)
 		removeLowArcs(graph, coverageCutoff);
@@ -635,7 +649,7 @@ int main(int argc, char **argv)
 	strcpy(graphFilename, directory);
 	strcat(graphFilename, "/contigs.fa");
 	sequenceLengths = getSequenceLengths(sequences, getWordLength(graph));
-	exportLongNodeSequences(graphFilename, graph, minContigKmerLength, sequences, sequenceLengths, coverageMask); 
+	exportLongNodeSequences(graphFilename, graph, minContigKmerLength, sequences, sequenceLengths, coverageMask);
 
 	if (exportAlignments) {
 		strcpy(graphFilename, directory);
@@ -663,9 +677,9 @@ int main(int argc, char **argv)
 	if (unusedReads)
 		exportUnusedReads(graph, sequences, minContigKmerLength, directory);
 
-	if (estimateCoverage) 
+	if (estimateCoverage)
 		velvetLog("Estimated Coverage = %f\n", expectedCoverage);
-	if (estimateCutoff) 
+	if (estimateCutoff)
 		velvetLog("Estimated Coverage cutoff = %f\n", coverageCutoff);
 
 	logFinalStats(graph, minContigKmerLength, directory);
@@ -673,25 +687,25 @@ int main(int argc, char **argv)
 	if (clean > 0) {
 		strcpy(graphFilename, directory);
 		strcat(graphFilename, "/Roadmaps");
-		remove(graphFilename);	
+		remove(graphFilename);
 
 		strcpy(graphFilename, directory);
 		strcat(graphFilename, "/LastGraph");
-		remove(graphFilename);	
-	} 
+		remove(graphFilename);
+	}
 
 	if (clean > 1) {
 		strcpy(graphFilename, directory);
 		strcat(graphFilename, "/Sequences");
-		remove(graphFilename);	
+		remove(graphFilename);
 
 		strcpy(graphFilename, directory);
 		strcat(graphFilename, "/Graph2");
-		remove(graphFilename);	
+		remove(graphFilename);
 
 		strcpy(graphFilename, directory);
 		strcat(graphFilename, "/Graph");
-		remove(graphFilename);	
+		remove(graphFilename);
 	}
 
 	free(sequenceLengths);

data/ext/src/src/utility.c

@@ -1,5 +1,5 @@
 /*
-Copyright 2009 John Marshall (jm18@sanger.ac.uk) 
+Copyright 2009 John Marshall (jm18@sanger.ac.uk)
 
     This file is part of Velvet.
 
@@ -89,14 +89,15 @@ void exitErrorf(int exitStatus, boolean showErrno, const char *format, ...)
        fprintf(stderr, "\n");
        va_end(args);
 
-#ifdef DEBUG 
+#ifdef DEBUG
 	abort();
-#endif 
+#endif
        exit(exitStatus);
 }
 
 void velvetLog(const char *format, ...)
 {
+  /*     Don't print anything as it interferes with code bound through bioruby-velvet_underground
   static boolean timeIsSet = false;
   static struct timeval tvStart;
   struct timeval tvNow;
@@ -117,14 +118,14 @@ void velvetLog(const char *format, ...)
   vprintf(format, args);
   va_end(args);
 
-#ifdef DEBUG 
+#ifdef DEBUG
   fflush(stdout);
-#endif
+#endif*/
 }
 
-void velvetFprintf(FILE * file, const char * format, ...) 
+void velvetFprintf(FILE * file, const char * format, ...)
 {
-	va_list args;
+  va_list args;
 
 	va_start(args, format);
 	if (vfprintf(file, format, args) < 0) {
@@ -132,9 +133,9 @@ void velvetFprintf(FILE * file, const char * format, ...)
 		       fprintf(stderr, "%s: ", programName);
 		fprintf(stderr, "Could not write into file\n");
 		va_end(args);
-#ifdef DEBUG 
+#ifdef DEBUG
 		abort();
-#endif 
+#endif
 		exit(EXIT_FAILURE);
 	}
 	va_end(args);

data/lib/bio-velvet_underground.rb

@@ -1,12 +1,56 @@
-# Please require your code below, respecting the naming conventions in the
-# bioruby directory tree.
-#
-# For example, say you have a plugin named bio-plugin, the only uncommented
-# line in this file would be 
-#
-#   require 'bio/bio-plugin/plugin'
-#
-# In this file only require other files. Avoid other source code.
-
-require 'bio-velvet_underground/velvet_underground.rb'
+require 'ffi'
+
+require 'bio-velvet_underground/constants'
+require 'bio-logger'
+
+Bio::Log::LoggerPlus.new('bio-velvet_underground')
+module Bio
+  module Velvet
+    module UndergroundLogging
+      def log
+        Bio::Log::LoggerPlus['bio-velvet_underground']
+      end
+    end
+
+    class Underground
+      extend FFI::Library
+      include Bio::Velvet::UndergroundLogging
+      def self.log
+        Bio::Log::LoggerPlus['bio-velvet_underground']
+      end
+
+      # Return the minimum kmer length greater than or equal to the given
+      # graph hash length e.g. 29 => 31, 31 => 31, 33 => 63.
+      def self.compilation_max_kmer(graph_hash_length)
+        max_kmers.select{|k| graph_hash_length<=k}.min
+      end
+
+      # Attach the correct shared velvet library with ffi. Options:
+      # :kmer: attach library with at least this much kmer length
+      def self.attach_shared_library(velvet_compilation_options={})
+        max_kmer_length = nil
+        given_kmer = velvet_compilation_options[:kmer]
+        if !given_kmer.nil?
+          max_kmer_length = compilation_max_kmer(given_kmer)
+          raise "No installed velvet library available for max kmer #{given_kmer}" if max_kmer_length.nil?
+        end
+        log.debug "Found max kmer length #{max_kmer_length} to load with the velvet library"
+
+        # Set the ffi library path to the correct velvet one
+        lib_location = self.library_location_of(max_kmer_length)
+        log.debug "Loading velvet underground FFI library #{lib_location}.."
+        ffi_lib lib_location
+        log.debug "Velvet library loaded."
+
+        attach_graph_functions
+        attach_binary_sequence_functions
+        attach_runner_functions
+      end
+    end
+  end
+end
+
+require 'bio-velvet_underground/binary_sequence_store'
+require 'bio-velvet_underground/graph'
+require 'bio-velvet_underground/runner'
 

data/lib/bio-velvet_underground/binary_sequence_store.rb

@@ -0,0 +1,86 @@
+class Bio::Velvet::Underground
+  class BinarySequenceStore
+    # Parse a CnyUnifiedSeq file in so that sequences can be accessed
+    def initialize(cny_unified_seq_file)
+      Bio::Velvet::Underground.attach_shared_library
+      readset_pointer = Bio::Velvet::Underground.importCnyReadSet cny_unified_seq_file
+      @readset = Bio::Velvet::Underground::ReadSet.new(readset_pointer)
+    end
+
+    # Return a sequence from the store given its read ID.
+    def [](sequence_id)
+      if sequence_id==0 or sequence_id > @readset[:readCount]
+        raise "Invalid sequence_id #{sequence_id}"
+      end
+
+      pointer = Bio::Velvet::Underground.getTightStringInArray(
+        @readset[:tSequences], sequence_id-1
+        )
+      Bio::Velvet::Underground.readTightString pointer
+    end
+
+    # Number of sequences in this store
+    def length
+      @readset[:readCount]
+    end
+
+    # Returns true if the sequence ID refers to the
+    # second in a pair of sequences.
+    def is_second_in_pair?(sequence_id)
+      if sequence_id==0 or sequence_id > @readset[:readCount]
+        raise "Invalid sequence_id #{sequence_id}"
+      end
+      Bio::Velvet::Underground.isSecondInPair @readset, sequence_id-1
+    end
+
+    # Returns the ID of the given sequence_id's pair
+    def pair_id(sequence_id)
+      if is_second_in_pair?(sequence_id)
+        sequence_id-1
+      else
+        sequence_id+1
+      end
+    end
+  end
+
+  private
+  # struct readSet_st {
+  #  char **sequences;
+  #  TightString *tSequences;
+  #  char **labels;
+  #  char *tSeqMem;
+  #  Quality **confidenceScores;
+  #  Probability **kmerProbabilities;
+  #  IDnum *mateReads;
+  #  Category *categories;
+  #  unsigned char *secondInPair;
+  #  IDnum readCount;
+  # };
+  class ReadSet < FFI::Struct
+    layout :sequences, :pointer, # char **sequences;
+    :tSequences, :pointer, # TightString *tSequences;
+    :labels, :pointer, # char **labels;
+    :tSeqMem, :pointer, # char *tSeqMem; #TODO: they don't really mean char* here - meant as an unsigned short?
+    :confidenceScores, :pointer, # Quality **confidenceScores;
+    :kmerProbabilities, :pointer, # Probability **kmerProbabilities;
+    :mateReads, :pointer, # IDnum *mateReads;
+    :categories, :pointer, # Category *categories;
+    :secondInPair, :pointer, # unsigned char *secondInPair;
+    :readCount, :int32 # IDnum readCount;
+  end
+
+  def self.attach_binary_sequence_functions
+    # ReadSet *importCnyReadSet(char *filename);
+    attach_function :importCnyReadSet, [:string], :pointer
+
+    # char *readTightString(TightString * tString); #tightString.h
+    attach_function :readTightString, [:pointer], :string
+
+    # TightString *getTightStringInArray(TightString * tString,
+    #			   IDnum	 position);
+    attach_function :getTightStringInArray, [:pointer, :int32], :pointer
+
+    # boolean isSecondInPair(ReadSet * reads, IDnum index);
+    attach_function :isSecondInPair, [:pointer, :int32], :bool
+  end
+end