obi import: added SINTAX format import and switch to version 3.0.1b18

python/obitools3/apps/optiongroups/__init__.py

@@ -57,6 +57,18 @@ def __addImportInputOption(optionManager):
                      const=b'rdp',
                      help="Input file is in RDP training set fasta format. If NCBI taxonomy provided with --taxonomy, taxid and scientific name will be added for each sequence.")
 
+    group.add_argument('--unite-input',
+                     action="store_const", dest="obi:inputformat",
+                     default=None,
+                     const=b'unite',
+                     help="Input file is in UNITE fasta format. If NCBI taxonomy provided with --taxonomy, taxid and scientific name will be added for each sequence.")
+
+    group.add_argument('--sintax-input',
+                     action="store_const", dest="obi:inputformat",
+                     default=None,
+                     const=b'sintax',
+                     help="Input file is in SINTAX fasta format. If NCBI taxonomy provided with --taxonomy, taxid and scientific name will be added for each sequence.")
+
     group.add_argument('--embl-input',
                      action="store_const", dest="obi:inputformat",
                      default=None,

python/obitools3/commands/import.pyx

@@ -2,6 +2,7 @@
 
 import sys
 import os
+import re
 
 from obitools3.apps.progress cimport ProgressBar  # @UnresolvedImport
 from obitools3.dms.view.view cimport View
@@ -108,6 +109,8 @@ def run(config):
     cdef   bint        NUC_SEQS_view
     cdef   bint        silva
     cdef   bint        rdp
+    cdef   bint        unite
+    cdef   bint        sintax
     cdef   int         nb_elts
     cdef   object      d
     cdef   View        view
@@ -233,16 +236,27 @@ def run(config):
         def_col = view[DEFINITION_COLUMN]
         seq_col = view[NUC_SEQUENCE_COLUMN]
 
-    # Prepare taxon scientific name and taxid refs if RDP or SILVA file
+    # Prepare taxon scientific name and taxid refs if RDP/SILVA/UNITE/SINTAX formats
     silva = False
     rdp = False
-    if 'inputformat' in config['obi'] and (config['obi']['inputformat'] == b"silva" or config['obi']['inputformat'] == b"rdp"):
+    unite = False
+    sintax=False
+    if 'inputformat' in config['obi'] and (config['obi']['inputformat'] == b"silva" or \
+                                           config['obi']['inputformat'] == b"rdp" or \
+                                           config['obi']['inputformat'] == b"unite" or \
+                                           config['obi']['inputformat'] == b"sintax"):
         #if taxo is None:
         #    raise Exception("A taxonomy (as built by 'obi import --taxdump') must be provided for SILVA and RDP files")
+        if config['obi']['inputformat'] == b"silva":
             silva = True
+        elif config['obi']['inputformat'] == b"rdp":
             rdp = True
-        if taxo is not None:
+        elif config['obi']['inputformat'] == b"unite":
+            unite = True
+        elif config['obi']['inputformat'] == b"sintax":
+            sintax = True
         sci_name_col = Column.new_column(view, SCIENTIFIC_NAME_COLUMN, OBI_STR)
+        if taxo is not None:
             taxid_col = Column.new_column(view, TAXID_COLUMN, OBI_INT)
 
     dcols = {}
@@ -349,11 +363,33 @@ def run(config):
                 if get_quality:
                     qual_col[i] = entry.quality
                     
-                # Parse taxon scientific name if RDP file
+                # Parse taxon scientific name if RDP or Silva or Unite file
-                if (rdp or silva) and taxo is not None:
+                if (rdp or silva or unite or sintax):
+                    if rdp or silva:
                         sci_names = entry.definition.split(b";")
+                        sci_name_col[i] = sci_names[-1]
+                    elif unite:
+                        sci_names = entry.id.split(b'|')[-1].split(b';')
+                        sci_name_col[i] = re.sub(b'[a-zA-Z]__', b'', sci_names[-1])
+                    elif sintax:
+                        reconstructed_line = entry.id+b' '+entry.definition[:-1]
+                        splitted_reconstructed_line = reconstructed_line.split(b';')
+                        taxa = splitted_reconstructed_line[1].split(b'=')[1]
+                        taxa = splitted_reconstructed_line[1].split(b',')
+                        sci_names = []
+                        for t in taxa:
+                            tf = t.split(b':')[1]
+                            sci_names.append(tf)
+                        sci_name_col[i] = sci_names[-1]
+                        id_col[i] = reconstructed_line.split(b';')[0]    
+                        def_col[i] = reconstructed_line
+                        
+                    # Fond taxid if taxonomy provided
+                    if taxo is not None :
                         for sci_name in reversed(sci_names):
-                        if sci_name.split()[0] != b'unidentified' and sci_name.split()[0] != b'uncultured' and sci_name.split()[0] != b'metagenome' :
+                            if unite:
+                                sci_name = re.sub(b'[a-zA-Z]__', b'', sci_name)
+                            if sci_name.split()[0] != b'unidentified' and sci_name.split()[0] != b'uncultured' and sci_name.split()[0] != b'metagenome':
                                 taxon = taxo.get_taxon_by_name(sci_name)
                                 if taxon is not None:
                                     sci_name_col[i] = taxon.name

python/obitools3/uri/decode.pyx

@@ -506,7 +506,7 @@ def open_uri(uri,
         if format is not None:
             if seqtype==b"nuc":
                 objclass = Nuc_Seq    # Nuc_Seq_Stored? TODO
-                if format==b"fasta" or format==b"silva" or format==b"rdp":
+                if format==b"fasta" or format==b"silva" or format==b"rdp" or format == b"unite" or format == b"sintax":
                     if input:
                         iseq = fastaNucIterator(file, 
                                                 skip=skip, 

python/obitools3/version.py

@@ -1,5 +1,5 @@
 major = 3
 minor = 0
-serial= '1b16'
+serial= '1b18'
 
 version ="%d.%d.%s" % (major,minor,serial)