Fix bug when VEP is run on an empty VCF file

Snakefile

@@ -14,7 +14,7 @@ localrules: create_summary, generate_report, package_results
 
 containers = {
     "debian": "docker://debian:buster-slim",
-    "fsnviz": "docker://quay.io/biocontainers/fsnviz:0.3.0--pyhdfd78af_5",
+    "crimson": "docker://quay.io/biocontainers/crimson:1.0.0--pyh5e36f6f_0",
     "hamlet-scripts": "docker://quay.io/redmar_van_den_berg/hamlet-scripts:0.3",
     "zip": "docker://quay.io/redmar_van_den_berg/zip:3.0"
 }
@@ -119,7 +119,7 @@ rule create_summary:
         run_name=RUN_NAME,
     output:
         js=OUTPUTS["summary"]
-    singularity: containers["fsnviz"]
+    singularity: containers["crimson"]
     shell:
         "python {input.scr}"
         " {input.idm}"

includes/snv-indels/Snakefile

@@ -225,6 +225,7 @@ rule annotate_vars:
         " --custom {input.ref_gonl},GONL,vcf,exact,0,AF"
         " --fork {threads} --offline --cache_version 97 --assembly GRCh38 --everything --merged"
         " --allele_number --stats_text --vcf --cache --force_overwrite"
+        " --format vcf"
         " --stats_file {output.stats} -o STDOUT | bgzip -c > {output.vcf}"
 
 rule extract_vars:

scripts/create_summary.py

@@ -3,6 +3,7 @@
 import json
 import csv
 from pathlib import Path
+from collections import defaultdict
 
 import click
 from crimson import picard, vep
@@ -119,16 +120,20 @@ def process_insert_stats(path):
 
 def process_var_stats(path):
     pd = vep.parse(path)
+
+    # If there are no variants, insert an empty defaultdict so that any queried
+    # value returns 0
+    if "Variants by chromosome" not in pd:
+        pd["Variants by chromosome"] = defaultdict(int)
+
     return {
-        "coding_consequences":
-        {k: v for k, v in pd["Coding consequences"].items()},
+        "coding_consequences": {k: v for k, v in pd["Coding consequences"].items()},
         "num_deletions": pd["Variant classes"].get("deletion", 0),
         "num_insertions": pd["Variant classes"].get("insertion", 0),
         "num_snvs": pd["Variant classes"].get("SNV", 0),
-        "per_chromosome":
-        {k: v
-         for k, v in pd["Variants by chromosome"].items()
-         if k in {str(i) for i in range(1, 23)}.union({"X", "Y", "MT"})},
+        "per_chromosome": {k: v
+                             for k, v in pd["Variants by chromosome"].items()
+                             if k in {str(i) for i in range(1, 23)}.union({"X", "Y", "MT"})},
         "polyphen": {
             "num_benign_variants": pd["PolyPhen summary"].get("benign", 0),
             "num_possibly_damaging_variants":