Merge pull request #385 from CGATOxford/{TS}-DebugCellTag

avoids keyError when generating random reads

tests/tests.yaml

@@ -217,7 +217,7 @@ dedup_single_tag_missing:
       stdin: chr19_tag_missing.bam
       outputs: [stdout]
       references: [single_tag_missing_py3.sam]
-      options: dedup -L test_missing.log --out-sam --random-seed=123456789 --method=directional --umi-tag=RX --extract-umi-method=tag
+      options: dedup -L test_missing.log --out-sam --random-seed=123456789 --method=directional --umi-tag=RX --extract-umi-method=tag --output-stats=stats_tag_missing
 
 dedup_single_gene_tag:
       stdin: chr19_gene_tags.bam
@@ -311,11 +311,11 @@ group_paired_use_unmapped:
       references: [paired_group_use_unmapped.sam]
       options: group -L test.log --out-sam --random-seed=123456789 --method=directional --no-sort-output --output-bam --unmapped=use --paired
 
-### Tests to implement ####
+# ### Tests to implement ####
 
-## dedup ##
-# mapping-quality (Need a BAM with meaningful MAPQ)
-# paired end (obv. need a paired end BAM)
-# read-length (ideally with sRNA-Seq BAM?!)
+# ## dedup ##
+# # mapping-quality (Need a BAM with meaningful MAPQ)
+# # paired end (obv. need a paired end BAM)
+# # read-length (ideally with sRNA-Seq BAM?!)
 
-# unpaired reads (Need a BAM with mixed paired and single end reads)
+# # unpaired reads (Need a BAM with mixed paired and single end reads)

umi_tools/umi_methods.py

@@ -149,6 +149,7 @@ def __init__(self, bamfile, chrom, barcode_getter):
         self.umis = collections.defaultdict(int)
         self.barcode_getter = barcode_getter
         self.random_fill_size = 100000  # Higher = faster, more memory
+        self.first_kerror = 1
         self.fill()
 
     def refill_random(self):
@@ -169,7 +170,10 @@ def fill(self):
             if read.is_read2:
                 continue
 
-            self.umis[self.barcode_getter(read)[0]] += 1
+            try:
+                self.umis[self.barcode_getter(read)[0]] += 1
+            except KeyError:
+                continue
 
         self.umis_counter = collections.Counter(self.umis)
         total_umis = sum(self.umis_counter.values())

umi_tools/version.py

@@ -1 +1 @@
-__version__ = "1.0.0"
+__version__ = "1.0.1"