Merge branch 'development'

docs/source/conf.py

@@ -26,7 +26,7 @@
 # The short X.Y version
 version = u'2.4'
 # The full version, including alpha/beta/rc tags
-release = u'2.4.1'
+release = u'2.4.2'
 
 
 # -- General configuration ---------------------------------------------------

setup.cfg

@@ -1,5 +1,5 @@
 [bumpversion]
-current_version = 2.4.1
+current_version = 2.4.2
 commit = True
 tag = True
 

setup.py

@@ -45,6 +45,6 @@
     test_suite='tests',
     tests_require=test_requirements,
     url='https://github.com/ratschlab/spladder',
-    version='2.4.1',
+    version='2.4.2',
     zip_safe=False,
 )

spladder/__init__.py

@@ -4,4 +4,4 @@
 
 __author__ = """Andre Kahles"""
 __email__ = 'andre.kahles@inf.ethz.ch'
-__version__ = '2.4.1'
+__version__ = '2.4.2'

spladder/count.py

@@ -71,9 +71,9 @@ def count_graph_coverage(genes, fn_bam=None, options=None, fn_out=None):
 
                 if options.sparse_bam and \
                   (fn_bam[f].endswith('npz') or \
-                   os.path.exists(re.sub(r'bam$', '', fn_bam[f]) + 'npz') or \
+                   os.path.exists(re.sub(r'[bB][aA][mM]|[cC][rR][aA][mM]$', '', fn_bam[f]) + 'npz') or \
                    fn_bam[f].endswith('hdf5') or \
-                   os.path.exists(re.sub(r'bam$', '', fn_bam[f]) + 'hdf5')):
+                   os.path.exists(re.sub(r'[bB][aA][mM]|[cC][rR][aA][mM]$', '', fn_bam[f]) + 'hdf5')):
                     ### make sure that we query the right contig from cache
                     assert(gg.chr == contig)
                     (tracks, intron_list) = add_reads_from_sparse_bam(gg, fn_bam[f], contig, options.confidence, types=['exon_track','intron_list'], filter=None, cache=bam_cache)
@@ -290,7 +290,7 @@ def collect_single_quantification_results(fname_out, sample_idxs, options):
                 h5fid.create_dataset(name='edges', data=CIN['edges'][:], chunks=True, compression='gzip', maxshape=(CIN['edges'].shape[0], None))
                 h5fid.create_dataset(name='segments', data=CIN['segments'][:], chunks=True, compression='gzip', maxshape=(CIN['segments'].shape[0], None))
                 h5fid.create_dataset(name='seg_pos', data=CIN['seg_pos'][:], chunks=True, compression='gzip', maxshape=(CIN['seg_pos'].shape[0], None))
-                h5fid.create_dataset(name='strains', data=CIN['strains'][:], chunks=True, compression='gzip', maxshape=(None,))
+                h5fid.create_dataset(name='strains', data=CIN['strains'][:], dtype='|S255', chunks=True, compression='gzip', maxshape=(None,))
             else:
                 appendToHDF5(h5fid, CIN['edges'][:], 'edges', faxis=1, daxis=1)
                 appendToHDF5(h5fid, CIN['segments'][:], 'segments', faxis=1, daxis=1)

spladder/helpers.py

@@ -100,7 +100,7 @@ def get_filename(which, options, sample_idx=None):
     if which in ['fn_count_in', 'fn_count_out']:
         if options.spladderfile == '-':
             if options.merge == 'single':
-                fname = os.path.join(options.outdir, 'spladder', 'genes_graph_conf%i.%s%s.pickle' % (options.confidence, options.samples[sample_idx], prune_tag))
+                fname = os.path.join(options.outdir, 'spladder', 'genes_graph_conf%i.single.%s%s.pickle' % (options.confidence, options.samples[sample_idx], prune_tag))
             else:
                 if (options.qmode == 'single' and which != 'fn_count_in') or (options.qmode == 'collect' and which == 'fn_count_in'):
                     fname = os.path.join(options.outdir, 'spladder', 'genes_graph_conf%i.%s.%s%s%s.pickle' % (options.confidence, options.merge, options.samples[sample_idx], prune_tag, validate_tag))

spladder/merge.py

@@ -9,6 +9,7 @@
     __package__ = 'modules'
 
 from .utils import *
+from .classes.segmentgraph import Segmentgraph
 from .count import count_graph_coverage_wrapper
 from .editgraph import filter_by_edgecount
 from . import rproc as rp
@@ -330,6 +331,8 @@ def merge_genes_by_splicegraph(options, merge_list=None, fn_out=None):
 
     for g in genes:
         g.label_alt()
+        if not g.segmentgraph.is_empty():
+            g.segmentgraph = Segmentgraph(g)
         g.to_sparse()
 
     ### re-sort genes by position - makes quantification more efficient

spladder/reads.py

@@ -282,7 +282,7 @@ def add_reads_from_sparse_bam(gg, fname, contig, conf, types=None, filter=None,
             if not filter is None:
                 IN = h5py.File(re.sub(r'[bB][aA][mM]|[cC][rR][aA][mM]$', '', fname) + 'conf_%i.' % conf + 'filt.' + 'hdf5', 'r')
             else:
-                IN = h5py.File(re.sub(r'[bB][aA][mM]|[cC][rR][aA][mM]$', '', fname) + 'hdf5')
+                IN = h5py.File(re.sub(r'[bB][aA][mM]|[cC][rR][aA][mM]$', '', fname) + 'hdf5', 'r')
 
         ### re-build sparse matrix
         cache['reads'] = scipy.sparse.coo_matrix((IN[contig + '_reads_dat'][:], (IN[contig + '_reads_row'][:], IN[contig + '_reads_col'][:])), shape=IN[contig + '_reads_shp'][:], dtype='uint32').tocsc()

spladder/settings.py

@@ -188,6 +188,8 @@ def parse_args(options, identity='main'):
         ### check existence of alignment files
         for fname in flatten(options.bam_fnames):
             if not os.path.isfile(fname):
+                if options.sparse_bam and os.path.isfile(re.sub(r'.[bB][aA][mM]$', '.hdf5', fname)):
+                    continue
                 print('ERROR: Input file %s can not be found\n\n' % fname, file=sys.stderr)
                 sys.exit(2)
             if not re.search(r'.[bB][aA][mM]$', fname) is None and not os.path.isfile(fname + '.bai'):