added metaSPAdes

assembly.py

@@ -354,6 +354,7 @@ def assemble_spades(
     filter_contigs=False,
     min_contig_len=0,
     kmer_sizes=(55,65),
+    spades_mode='rna',
     n_reads=10000000,
     outReads=None,
     always_succeed=False,
@@ -378,7 +379,7 @@ def assemble_spades(
             tools.spades.SpadesTool().assemble(reads_fwd=reads_fwd, reads_bwd=reads_bwd, reads_unpaired=reads_unpaired,
                                                contigs_untrusted=contigs_untrusted, contigs_trusted=contigs_trusted,
                                                contigs_out=out_fasta, filter_contigs=filter_contigs,
-                                               min_contig_len=min_contig_len,
+                                               min_contig_len=min_contig_len, spades_mode=spades_mode,
                                                kmer_sizes=kmer_sizes, always_succeed=always_succeed, max_kmer_sizes=max_kmer_sizes,
                                                spades_opts=spades_opts, mem_limit_gb=mem_limit_gb,
                                                threads=threads)
@@ -403,6 +404,8 @@ def parser_assemble_spades(parser=argparse.ArgumentParser()):
     parser.add_argument('--outReads', default=None, help='Save the trimmomatic/prinseq/subsamp reads to a BAM file')
     parser.add_argument('--filterContigs', dest='filter_contigs', default=False, action='store_true', 
                         help='only output contigs SPAdes is sure of (drop lesser-quality contigs from output)')
+    parser.add_argument('--spadesMode', dest='spades_mode', choices=('rna', 'meta'), default='meta',
+                        help='which SPAdes mode to use (rnaSPAdes or metaSPAdes)')
     parser.add_argument('--alwaysSucceed', dest='always_succeed', default=False, action='store_true',
                         help='if assembly fails for any reason, output an empty contigs file, rather than failing with '
                         'an error code')

pipes/WDL/workflows/tasks/tasks_assembly.wdl

@@ -5,6 +5,7 @@ task assemble {
     Int?    trinity_n_reads=250000
     Int?    spades_n_reads=10000000
     Int?    spades_min_contig_len=0
+    String? spades_mode="rna"
 
     String? assembler="trinity"  # trinity, spades, or trinity-spades
     Boolean? always_succeed=false
@@ -39,6 +40,7 @@ task assemble {
             ${'--nReads=' + spades_n_reads} \
 	    ${true="--alwaysSucceed" false="" always_succeed} \
             ${'--minContigLen=' + spades_min_contig_len} \
+            ${'--spadesMode=' + spades_mode} \
             --memLimitGb $mem_in_gb \
             --outReads=${sample_name}.subsamp.bam \
             --loglevel=DEBUG
@@ -61,6 +63,7 @@ task assemble {
             ${'--nReads=' + spades_n_reads} \
      	    ${true='--alwaysSucceed' false='' always_succeed} \
             ${'--minContigLen=' + spades_min_contig_len} \
+            ${'--spadesMode=' + spades_mode} \
             --memLimitGb $mem_in_gb \
             --loglevel=DEBUG
 

tools/spades.py

@@ -43,7 +43,8 @@ def execute(self, args):    # pylint: disable=W0221
 
     def assemble(self, reads_fwd, reads_bwd, contigs_out, reads_unpaired=None, contigs_trusted=None,
                  contigs_untrusted=None, kmer_sizes=(55,65), always_succeed=False, max_kmer_sizes=1, 
-                 filter_contigs=False, min_contig_len=0, mem_limit_gb=8, threads=None, spades_opts=''):
+                 filter_contigs=False, min_contig_len=0, mem_limit_gb=8, threads=None, spades_opts='',
+                 spades_mode='rna'):
         '''Assemble contigs from RNA-seq reads and (optionally) pre-existing contigs.
 
         Inputs:
@@ -63,6 +64,7 @@ def assemble(self, reads_fwd, reads_bwd, contigs_out, reads_unpaired=None, conti
             mem_limit_gb: max memory to use, in gigabytes
             threads: number of threads to use
             spades_opts: additional options to pass to spades
+            spades_mode: whether to use rnaSPAdes (spades_mode=="rna") or metaSPAdes (spades_mode=="meta")
         Outputs:
             contigs_out: assembled contigs in fasta format.  Note that, since we use the
                 RNA-seq assembly mode, for some genome regions we may get several contigs
@@ -89,14 +91,20 @@ def assemble(self, reads_fwd, reads_bwd, contigs_out, reads_unpaired=None, conti
                     if reads_fwd and reads_bwd and os.path.getsize(reads_fwd) > 0 and os.path.getsize(reads_bwd) > 0:
                         args += ['-1', reads_fwd, '-2', reads_bwd ]
                     if reads_unpaired and os.path.getsize(reads_unpaired) > 0:
-                        args += [ '--s1', reads_unpaired ]
+                        if spades_mode == 'meta':
+                            log.warning('SPAdes: ignoring unpaired reads in metaSPAdes mode')
+                        else:
+                            args += [ '--s1', reads_unpaired ]
+
                     if contigs_trusted: args += [ '--trusted-contigs', contigs_trusted ]
                     if contigs_untrusted: args += [ '--untrusted-contigs', contigs_untrusted ]
                     if kmer_size: args += [ '-k', kmer_size ]
                     if spades_opts: args += shlex.split(spades_opts)
-                    args += [ '--rna', '-m' + str(mem_limit_gb), '-t', str(threads), '-o', spades_dir ]
+                    args += [ '--' + spades_mode, '-m' + str(mem_limit_gb), '-t', str(threads), '-o', spades_dir ]
 
-                    transcripts_fname = os.path.join(spades_dir, ('hard_filtered_' if filter_contigs else '') + 'transcripts.fasta')
+                    transcripts_fname = os.path.join(spades_dir,
+                                                     ('hard_filtered_' if filter_contigs else '') + 'transcripts.fasta') \
+                                        if spades_mode == 'rna' else 'contigs.fasta'
 
                     try:
                         self.execute(args=args)