added some filter settings to CLI

Author: iskandr

script/isovar-translate-variants.py

@@ -20,6 +20,7 @@
 
 import varcode
 import skbio
+import numpy as np
 from pysam import AlignmentFile
 
 from isovar import gather_variant_reads, sequence_counts
@@ -39,13 +40,19 @@
     default=None)
 
 parser.add_argument(
-    "--min-count", type=int, default=3)
+    "--min-read-count",
+    type=int,
+    default=3)
 
 parser.add_argument(
-    "--context-size",
-    default=45,
+    "--sequence-length",
+    default=105,
     type=int)
 
+parser.add_argument(
+    "--max-sequences-per-variant",
+    type=int,
+    default=5)
 
 if __name__ == "__main__":
     args = parser.parse_args()
@@ -54,31 +61,42 @@
     samfile = AlignmentFile(args.bam)
 
     for variant in variants:
+        print(variant)
         variant_reads = gather_variant_reads(
             samfile=samfile,
             chromosome="chr" + variant.contig,
             base1_location=variant.start,
             ref=variant.ref,
             alt=variant.alt)
-        if len(variant_reads) < args.min_count:
+        if len(variant_reads) < args.min_read_count:
             continue
+
+        # the number of context nucleotides on either side of the variant
+        # is half the desired length (minus the number of variant nucleotides)
+        context_size = int(
+            np.ceil((args.sequence_length - len(variant.alt)) / 2.0))
         sequence_count_info = sequence_counts(
-            variant_reads, context_size=args.context_size)
-        for ((prefix, suffix), count) in sorted(
+            variant_reads,
+            context_size=context_size)
+        for i, ((prefix, suffix), count) in enumerate(sorted(
                 sequence_count_info.full_read_counts.items(),
-                key=lambda x: -x[1]):
-            if count < args.min_count:
+                key=lambda x: -x[1])):
+            if i >= args.max_sequences_per_variant:
                 break
 
-            variant = sequence_count_info.variant_nucleotides
+            if count < args.min_read_count:
+                break
+
+            variant_seq = sequence_count_info.variant_nucleotides
+
             print("\t%s_%s_%s: %d" % (
                 prefix,
-                variant,
+                variant_seq,
                 suffix,
                 count))
 
             # translate in three reading frames:
-            seq = "%s%s%s" % (prefix, variant, suffix)
+            seq = "%s%s%s" % (prefix, variant_seq, suffix)
             for offset in range(3):
                 dna = skbio.DNA(seq[offset:])
                 print("\t\tframe=%d: %s" % (offset, dna.translate()))