fixed deprecation warnings

Author: akahles

spladder/alt_splice/events.py

@@ -107,7 +107,7 @@ def make_unique_by_event(event_list):
             last_kept = i
 
     print('events dropped: %i' % len(rm_idx))
-    keep_idx = np.where(~np.in1d(np.arange(event_list.shape[0]), rm_idx))[0]
+    keep_idx = np.where(~np.isin(np.arange(event_list.shape[0]), rm_idx))[0]
     event_list = event_list[keep_idx]
 
     return event_list
@@ -158,7 +158,7 @@ def curate_alt_prime(event_list, options):
 
     ### remove events with non-overlapping alt_exons
     if len(rm_idx) > 0:
-        keep_idx = np.where(~np.in1d(np.arange(event_list.shape[0]), rm_idx))[0]
+        keep_idx = np.where(~np.isin(np.arange(event_list.shape[0]), rm_idx))[0]
         event_list = event_list[keep_idx]
 
     print('Corrected %i events' % corr_count)

spladder/classes/event.py

@@ -51,7 +51,7 @@ def get_exon_coordinate_strings(self):
         sidx = np.argsort([int(_.split('-')[0]) for _ in _iso_both])
         _iso_both = _iso_both[sidx]
 
-        _usage = (~(np.in1d(_iso_both, _iso1) & np.in1d(_iso_both, _iso2))).astype('int')
+        _usage = (~(np.isin(_iso_both, _iso1) & np.isin(_iso_both, _iso2))).astype('int')
 
         return ':'.join(_iso_both), ':'.join(_usage.astype('str'))
 

spladder/classes/gene.py

@@ -84,7 +84,7 @@ def label_alt(self):
             ### handle start terminal exons -> no incoming edges
             if not np.any(edges[:i, i]):
                 ### find other exons with same end
-                idx = np.where(~np.in1d(np.where(vertices[1, i] == vertices[1, :])[0], i))[0]
+                idx = np.where(~np.isin(np.where(vertices[1, i] == vertices[1, :])[0], i))[0]
                 if idx.shape[0] > 0:
                     is_simple = True
                     for j in idx:
@@ -105,7 +105,7 @@ def label_alt(self):
             ### handle end terminal exons -> no outgoing edges
             if not np.any(edges[i, i+1:]):
                 ### find other exons with the same start
-                idx = np.where(~np.in1d(np.where(vertices[0, i] == vertices[0, :])[0], i))[0]
+                idx = np.where(~np.isin(np.where(vertices[0, i] == vertices[0, :])[0], i))[0]
                 if idx.shape[0] > 0: 
                     is_simple = True
                     for j in idx:
@@ -125,7 +125,7 @@ def label_alt(self):
                     term_alt_idx.append(i)
 
         ### further only consider exons that are neither init_alt nor term_alt
-        take_idx = np.where(~np.in1d(np.arange(num_exons), [init_alt_idx + term_alt_idx]))[0]
+        take_idx = np.where(~np.isin(np.arange(num_exons), [init_alt_idx + term_alt_idx]))[0]
         if take_idx.shape[0] > 0:
             vertices = self.splicegraph.vertices[:, take_idx]
             edges = self.splicegraph.edges[take_idx, :][:, take_idx]

spladder/classes/splicegraph.py

@@ -284,7 +284,7 @@ def uniquify(self):
                 self.edges[j, :] = self.edges[j-1, :] | self.edges[j, :]
                 rm_idx.append(j - 1)
 
-        keep_idx = np.where(~np.in1d(np.array(np.arange(self.vertices.shape[1])), rm_idx))[0]
+        keep_idx = np.where(~np.isin(np.array(np.arange(self.vertices.shape[1])), rm_idx))[0]
         self.vertices = self.vertices[:, keep_idx]
         self.edges = self.edges[keep_idx, :][:, keep_idx]
         self.terminals = self.terminals[:, keep_idx]

spladder/editgraph.py

@@ -64,10 +64,10 @@ def remove_short_exons(genes, options):
                         short_exon_skipped += 1
             j += 1
 
-        keep_idx = np.where(~np.in1d(np.array(np.arange(genes[i].splicegraph.vertices.shape[1])), exons_remove_idx))[0]
+        keep_idx = np.where(~np.isin(np.array(np.arange(genes[i].splicegraph.vertices.shape[1])), exons_remove_idx))[0]
         genes[i].splicegraph.subset(keep_idx)
 
-    keep_idx = np.where(~np.in1d(np.arange(len(genes)), rm_idx))[0]
+    keep_idx = np.where(~np.isin(np.arange(len(genes)), rm_idx))[0]
     genes = genes[keep_idx]
 
     if options.verbose:
@@ -165,7 +165,7 @@ def reduce_splice_graph(genes):
                             ###              <--- test_exon ---->           <---- test_exon ---->>>>>>>>>>>>
                             if ((vertices[1, exon_idx] >= vertices[0, test_exon_idx]) and (vertices[0, exon_idx] <= vertices[0, test_exon_idx])) or \
                                ((vertices[1, test_exon_idx] >= vertices[0, exon_idx]) and (vertices[0, test_exon_idx] <= vertices[0, exon_idx])) and \
-                               (np.sum(np.in1d(np.arange(min(vertices[0, exon_idx], vertices[0, test_exon_idx]), max(vertices[1, exon_idx], vertices[1, test_exon_idx])), intron_loc)) == 0):
+                               (np.sum(np.isin(np.arange(min(vertices[0, exon_idx], vertices[0, test_exon_idx]), max(vertices[1, exon_idx], vertices[1, test_exon_idx])), intron_loc)) == 0):
 
                                 ### merge exons if they overlap and they do not span any intronic position
                                 vertices[0, exon_idx] = min(vertices[0, exon_idx], vertices[0, test_exon_idx])
@@ -184,7 +184,7 @@ def reduce_splice_graph(genes):
                             ###   ----- exon -----<
                             ###   --- test_exon --<
                             if (vertices[1, exon_idx] == vertices[1, test_exon_idx]) and \
-                               (np.sum(np.in1d(np.arange(min(vertices[0, exon_idx], vertices[0, test_exon_idx]), vertices[1, exon_idx]), intron_loc)) == 0):
+                               (np.sum(np.isin(np.arange(min(vertices[0, exon_idx], vertices[0, test_exon_idx]), vertices[1, exon_idx]), intron_loc)) == 0):
 
                                 ### merge exons if they share the same right boundary and do not span intronic positions
                                 vertices[0, exon_idx] = min(vertices[0, exon_idx], vertices[0, test_exon_idx])
@@ -204,7 +204,7 @@ def reduce_splice_graph(genes):
                             ###   >---- exon ------
                             ###   >-- test_exon ---
                             if (vertices[0, exon_idx] == vertices[0, test_exon_idx]) and \
-                               (np.sum(np.in1d(np.arange(vertices[0, exon_idx], max(vertices[1, exon_idx], vertices[1, test_exon_idx])), intron_loc)) == 0):
+                               (np.sum(np.isin(np.arange(vertices[0, exon_idx], max(vertices[1, exon_idx], vertices[1, test_exon_idx])), intron_loc)) == 0):
 
                                 ### merge exons if they share the same left boundary and do not span intronic positions
                                 vertices[1, exon_idx] = max(vertices[1, exon_idx], vertices[1, test_exon_idx])
@@ -257,8 +257,8 @@ def filter_by_edgecount(genes, options):
         genes[i].splicegraph.edges = (genes[i].edge_count >= options.sg_min_edge_count).astype('int')
         ### remove all exons that have no incoming or outgoing edges (caused by validation, keep single exon transcripts that occured before)
         k_idx2 = np.where(genes[i].splicegraph.edges.sum(axis = 1) > 0)[0]
-        #rm_idx = np.where(~np.in1d(k_idx2, k_idx))[0]
-        #keep_idx = np.where(~np.in1d(np.arange(genes[i].splicegraph.edges.shape[0]), rm_idx))[0]
+        #rm_idx = np.where(~np.isin(k_idx2, k_idx))[0]
+        #keep_idx = np.where(~np.isin(np.arange(genes[i].splicegraph.edges.shape[0]), rm_idx))[0]
         keep_idx = np.union1d(k_idx, k_idx2).astype('int')
         if keep_idx.shape[0] > 0:
             genes[i].splicegraph.subset(keep_idx)
@@ -279,7 +279,7 @@ def insert_intron_retentions(genes, bam_fnames, options):
 
     ### ignore contigs not present in bam files 
     # TODO
-    #keepidx = np.where(np.in1d(np.array([options.chrm_lookup[x.chr] for x in genes[chunk_idx]]), np.array([x.chr_num for x in regions])))[0]
+    #keepidx = np.where(np.isin(np.array([options.chrm_lookup[x.chr] for x in genes[chunk_idx]]), np.array([x.chr_num for x in regions])))[0]
     #genes = genes[keepidx]
 
     c = 0 
@@ -509,7 +509,6 @@ def insert_intron_edges(genes, bam_fnames, options):
                                                                                                          genes[i].splicegraph.vertices[1, -2], genes[i].splicegraph.vertices[0, -1], 
                                                                                                          genes[i].splicegraph.vertices[1, -1]), file=fd_log)
                         intron_used = True
-
                 if not intron_used:
                     unused_introns.append(j)
                 continue # with next intron
@@ -672,6 +671,9 @@ def insert_intron_edges(genes, bam_fnames, options):
                                         tracks += tmp_
                             tracks = np.asarray(tracks)
                         else:
+                            #if i == 11:
+                            #    import pdb
+                            #    pdb.set_trace()
                             tracks = add_reads_from_bam(np.array([gg], dtype='object'), bam_fnames, ['exon_track'], options.read_filter, options.var_aware, options.primary_only, options.ignore_mismatches, mm_tag=options.mm_tag, cram_ref=options.ref_genome)
 
                         ### TODO: make configurable
@@ -835,7 +837,7 @@ def insert_cassette_exons(genes, bam_fnames, options):
 
     ### ignore contigs not present in bam files 
     # TODO TODO
-    #keepidx = np.where(np.in1d(np.array([CFG['chrm_lookup'][x.chr] for x in genes]), np.array([x.chr_num for x in regions])))[0]
+    #keepidx = np.where(np.isin(np.array([CFG['chrm_lookup'][x.chr] for x in genes]), np.array([x.chr_num for x in regions])))[0]
     #genes = genes[keepidx]
 
     c = 0

spladder/helpers.py

@@ -35,7 +35,7 @@ def make_introns_feasible(introns, genes, bam_fnames, options):
         tmp2 = np.array([x.shape[0] for x in introns[:, 1]])
 
         still_unfeas = np.where((tmp1 > 200) | (tmp2 > 200))[0]
-        idx = np.where(~np.in1d(unfeas, still_unfeas))[0]
+        idx = np.where(~np.isin(unfeas, still_unfeas))[0]
 
         for i in unfeas[idx]:
             print('[feasibility] set criteria for gene %s to: min_ex %i, min_conf %i, max_mism %i' % (genes[i].name, options.read_filter['exon_len'], options.read_filter['mincount'], options.read_filter['mismatch']), file=fd_log)
@@ -113,7 +113,7 @@ def filter_introns(introns, genes, options):
             k_idx = []
             cnt_tot += introns[i, si].shape[0]
             for j in range(introns[i, si].shape[0]):
-                if len(gene_trees[(s, genes[i].chr)].overlap(introns[i, si][j, 0] - offset, introns[i, si][j, 1] + offset)) == 1:
+                if len(gene_trees[(s, genes[i].chr)].overlap(int(introns[i, si][j, 0]) - offset, int(introns[i, si][j, 1]) + offset)) == 1:
                     k_idx.append(j)
             if len(k_idx) < introns[i, si].shape[0]:
                 cnt_rem += (introns[i, si].shape[0] - len(k_idx))
@@ -206,13 +206,10 @@ def log_progress(idx, total, bins=50):
     sys.stdout.flush()
 
 def codeUTF8(s):
-    return s.view(np.chararray).encode('utf-8')
+    return s.astype('bytes')
 
 def decodeUTF8(s):
-    return s.view(np.chararray).decode('utf-8')
-
-def isUTF8(s):
-    return hasattr(s.view(np.chararray), 'decode')
+    return s.astype('str')
 
 def rev_comp(s):
     R = {'A':'T', 'T':'A', 'a':'t', 't':'a',

spladder/helpers_viz.py

@@ -160,7 +160,7 @@ def get_conf_events(options, gid):
         IN = h5py.File(os.path.join(options.outdir, 'merge_graphs_%s_C%i.counts.hdf5' % (event_type, options.confidence)), 'r')
         if 'conf_idx' in IN and IN['conf_idx'].shape[0] > 0:
             conf_idx = IN['conf_idx'][:]
-            k_idx = sp.where(sp.in1d(IN['gene_idx'][:][conf_idx], gid))[0]
+            k_idx = sp.where(sp.isin(IN['gene_idx'][:][conf_idx], gid))[0]
             if k_idx.shape[0] > 0:
                 event_info.extend([[event_type, x] for x in conf_idx[k_idx]])
         IN.close()

spladder/merge.py

@@ -36,7 +36,7 @@ def merge_duplicate_exons(genes, options):
             if remove[j]:
                 continue
             idx = np.where((exons[0, j] == exons[0, :]) & (exons[1, j] == exons[1, :]))[0]
-            idx = idx[np.where(~np.in1d(idx, j))[0]]
+            idx = idx[np.where(~np.isin(idx, j))[0]]
             if idx.shape[0] > 0:
                 remove[idx] = True
                 for k in idx:

spladder/reads.py

@@ -7,6 +7,8 @@
 import h5py
 import uuid
 
+from collections import defaultdict
+
 if __package__ is None:
     __package__ = 'modules'
 
@@ -27,8 +29,8 @@ def get_reads(fname, chr_name, start, stop, strand=None, filter=None, mapped=Tru
     j = []
 
     read_cnt = 0
-    introns_p = dict()
-    introns_m = dict()
+    introns_p = defaultdict(int)
+    introns_m = defaultdict(int)
 
     if collapse:
         read_matrix = np.zeros((1, stop - start), dtype='int')
@@ -67,25 +69,26 @@ def get_reads(fname, chr_name, start, stop, strand=None, filter=None, mapped=Tru
             #   6 / P - padding
             #   7 / = - sequence match
             #   8 / X - sequence mismatch
-            p = read.pos 
+            p = int(read.pos)
             for o in read.cigar:
                 if o[0] == 3:
                     if is_minus:
-                        try:
-                            introns_m[(p, p + o[1])] += 1
-                        except KeyError:
-                            introns_m[(p, p + o[1])] = 1
+                        introns_m[(p, p + o[1])] += 1
                     else:
-                        try:
-                            introns_p[(p, p + o[1])] += 1
-                        except KeyError:
-                            introns_p[(p, p + o[1])] = 1
+                        introns_p[(p, p + o[1])] += 1
                 if o[0] in [0, 2, 7, 8]:
-                    _start = int(max(p-start, 0))
-                    _stop = int(min(p + o[1] - start, stop - start))
+                    if p + o[1] <= start:
+                        p += o[1]
+                        continue
+                    if p < start:
+                        _start = 0
+                        _len = int(o[1]) - int(start) + p
+                    else:
+                        _start = p - int(start)
+                        _len = int(o[1])
+                    _stop = min(_start + _len, int(stop))
                     if _stop < 0 or _start > length:
-                        if o[0] in [0, 2, 3, 7, 8]:
-                            p += o[1]
+                        p += o[1]
                         continue
                     if collapse:
                         read_matrix[0, _start:_stop] += 1
@@ -423,7 +426,7 @@ def get_intron_list(genes, bam_fnames, options):
     strands = ['+', '-']
 
     ### ignore contigs not present in bam files 
-    keepidx = np.where(np.in1d(np.array([options.chrm_lookup[x.chr] for x in genes]), np.array([x.chr_num for x in regions])))[0]
+    keepidx = np.where(np.isin(np.array([options.chrm_lookup[x.chr] for x in genes]), np.array([x.chr_num for x in regions])))[0]
     genes = genes[keepidx]
 
     c = 0
@@ -446,6 +449,7 @@ def get_intron_list(genes, bam_fnames, options):
 
                 gg = np.array([copy.copy(genes[i])], dtype='object')
                 assert(gg[0].strand == s)
+                # TODO intron window is hard coded --> make configurable
                 gg[0].start = max(gg[0].start - 5000, 1)
                 gg[0].stop = gg[0].stop + 5000
                 assert(gg[0].chr == contig)

spladder/spladder_viz.py

@@ -375,7 +375,7 @@ def spladder_viz(options):
                     if len(data_track.samples) > 0:
                         seg_sample_idx = []
                         for track_samples in data_track.samples:
-                            seg_sample_idx.append(np.where(np.in1d(samples, track_samples))[0])
+                            seg_sample_idx.append(np.where(np.isin(samples, track_samples))[0])
                             cov_from_segments(genes[g], segments, edges, edge_idx, size_factors, ax, xlim=plotrange, log=options.log, grid=True, order='C', sample_idx=seg_sample_idx)
                 ax.get_yaxis().set_label_coords(1.02,0.5)
                 ax.set_ylabel('segment counts')