Merge pull request #232 from pachterlab/devel

Devel

kb_python/count.py

@@ -204,6 +204,8 @@ def kallisto_bus(
 
     if technology.upper() in ('BULK', 'SMARTSEQ3'):
         results['saved_index'] = os.path.join(out_dir, SAVED_INDEX_FILENAME)
+        if os.path.exists(results['saved_index']):
+            os.remove(results['saved_index'])  # TODO: Fix this in kallisto?
     return results
 
 

kb_python/main.py

@@ -636,7 +636,7 @@ def parse_count(
             parser.error(
                 f'Technology `{args.x}` can not be used with workflow {args.workflow}.'
             )
-        if args.sum is not None:
+        if args.sum != "none":
             parser.error('--sum incompatible with lamanno/nucleus')
         if args.x.upper() == 'SMARTSEQ3':
             from .count import count_velocity_smartseq3
@@ -1036,8 +1036,8 @@ def setup_ref_args(
         '--workflow',
         metavar='{standard,nac,kite,custom}',
         help=(
-            'Type of workflow to prepare files for. '
-            'Use `nac` for RNA velocity or single-nucleus RNA-seq reads. '
+            'The type of index to create. '
+            'Use `nac` for an index type that can quantify nascent and mature RNA. '
             'Use `custom` for indexing targets directly. '
             'Use `kite` for feature barcoding. (default: standard)'
         ),
@@ -1253,7 +1253,7 @@ def setup_count_args(
         metavar='{standard,nac,kite,kite:10xFB}',
         help=(
             'Type of workflow. '
-            'Use `nac` for RNA velocity or single-nucleus RNA-seq reads. '
+            'Use `nac` to specify a nac index for producing mature/nascent/ambiguous matrices. '
             'Use `kite` for feature barcoding. '
             'Use `kite:10xFB` for 10x Genomics Feature Barcoding technology. '
             '(default: standard)'
@@ -1301,14 +1301,14 @@ def setup_count_args(
     required_nac.add_argument(
         '-c1',
         metavar='T2C',
-        help='Path to cDNA transcripts-to-capture',
+        help='Path to mature transcripts-to-capture',
         type=str,
         required=workflow in {'nac'}
     )
     required_nac.add_argument(
         '-c2',
         metavar='T2C',
-        help='Path to intron transcripts-to-captured',
+        help='Path to nascent transcripts-to-captured',
         type=str,
         required=workflow in {'nac'}
     )

kb_python/ref.py

@@ -282,7 +282,12 @@ def kallisto_index(
         command += ['--d-list-overhang', dlist_overhang]
     if temp_dir != 'tmp':
         command += ['-T', temp_dir]
-    command += [fasta_path]
+    if ',' in fasta_path:
+        fasta_paths = fasta_path.split(',')
+        for fp in fasta_paths:
+            command += [fp]
+    else:
+        command += [fasta_path]
     run_executable(command)
     return {'index': index_path}
 
@@ -303,11 +308,15 @@ def get_dlist_fasta(fasta_path: str = None, temp_dir: str = 'tmp') -> str:
     if "://" not in fasta_path:  # Not a URL
         return fasta_path
     new_fasta_path = get_temporary_filename(temp_dir)
+    fasta_path_array = [fasta_path]
+    if fasta_path.count("://") > 1:
+        fasta_path_array = fasta_path.split(",")
     logger.info(f'Extracting {fasta_path} into {new_fasta_path}')
-    with ngs.fasta.Fasta(fasta_path, 'r') as f_in:
-        with ngs.fasta.Fasta(new_fasta_path, 'w') as f_out:
-            for entry in f_in:
-                f_out.write(entry)
+    with ngs.fasta.Fasta(new_fasta_path, 'w') as f_out:
+        for fp in fasta_path_array:
+            with ngs.fasta.Fasta(fp, 'r') as f_in:
+                for entry in f_in:
+                    f_out.write(entry)
     return new_fasta_path
 
 
@@ -803,7 +812,7 @@ def ref_custom(
 
     if not glob.glob(f'{index_path}*') or overwrite:
         index_result = kallisto_index(
-            ' '.join(fasta_paths),
+            ','.join(fasta_paths),
             index_path,
             k=k or 31,
             threads=threads,

kb_python/utils.py

@@ -711,8 +711,8 @@ def overlay_anndatas(
     spliced_intersection = adata_spliced[obs_idx][:, var_idx]
     unspliced_intersection = adata_unspliced[obs_idx][:, var_idx]
     a_layers = {
-        'spliced': spliced_intersection.X,
-        'unspliced': unspliced_intersection.X
+        'mature': spliced_intersection.X,
+        'nascent': unspliced_intersection.X
     }
     ambiguous_intersection = None
     if adata_ambiguous is not None:

tests/test_utils.py

@@ -227,7 +227,7 @@ def test_overlay_anndatas(self):
             self.unspliced_genes_path
         )
         adata = utils.overlay_anndatas(adata_spliced, adata_unspliced)
-        self.assertEqual({'spliced', 'unspliced'}, set(adata.layers.keys()))
+        self.assertEqual({'mature', 'nascent'}, set(adata.layers.keys()))
 
     def test_sum_anndatas(self):
         adata_spliced = utils.import_matrix_as_anndata(