change --ont defaults and illumina error rate [Mykrobe-tools#133 & My…

…krobe-tools#134]

CHANGELOG.md

@@ -19,6 +19,9 @@ this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.htm
 
 ### Changed
 - Default kmer size (21) made consistent across all subcommands [[#141][141]]
+- Nanopore preset `--ont` defaults to an expected error rate of 0.08 and ploidy
+  'haploid' [[#134][134]]
+- Default Illumina expected error rate change from 0.05 to 0.001 [[#133][133]]
 
 ### Fixed
 - Improved error messaging when an X amino acid resolves to a mutation already present
@@ -63,6 +66,8 @@ this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.htm
 [119]: https://github.com/Mykrobe-tools/mykrobe/issues/119
 [123]: https://github.com/Mykrobe-tools/mykrobe/issues/123
 [127]: https://github.com/Mykrobe-tools/mykrobe/issues/127
+[133]: https://github.com/Mykrobe-tools/mykrobe/issues/133
+[134]: https://github.com/Mykrobe-tools/mykrobe/issues/134
 [140]: https://github.com/Mykrobe-tools/mykrobe/issues/140
 [141]: https://github.com/Mykrobe-tools/mykrobe/issues/141
 [Unreleased]: https://github.com/Mykrobe-tools/mykrobe/compare/v0.10.0...HEAD

src/mykrobe/base.py

@@ -3,7 +3,7 @@
 import argparse
 import os
 
-from mykrobe import K
+from mykrobe import K, ONT_E_RATE, ONT_PLOIDY, ILLUMINA_E_RATE
 
 sequence_or_graph_parser_mixin = argparse.ArgumentParser(add_help=False)
 sequence_or_graph_parser_mixin.add_argument(
@@ -75,7 +75,7 @@
 genotyping_mixin.add_argument(
     "--ont",
     action="store_true",
-    help="Set default for ONT data. Sets expected_error_rate to 0.15 and to haploid",
+    help=f"Set defaults for ONT data. Sets `-e {ONT_E_RATE} --ploidy {ONT_PLOIDY}`",
 )
 genotyping_mixin.add_argument(
     "--guess_sequence_method",
@@ -123,8 +123,8 @@
 genotyping_mixin.add_argument(
     "-e",
     "--expected_error_rate",
-    help="Expected sequencing error rate (default: %(default).2f)",
-    default=0.05,
+    help="Expected sequencing error rate (default: %(default).3f)",
+    default=ILLUMINA_E_RATE,
     type=float,
 )
 genotyping_mixin.add_argument(

src/mykrobe/cmds/amr.py

@@ -4,6 +4,8 @@
 import logging
 import tempfile
 
+from mykrobe import ONT_E_RATE, ONT_PLOIDY
+
 logger = logging.getLogger(__name__)
 
 import json
@@ -342,24 +344,10 @@ def run(parser, args):
             args.ont = True
 
         if args.ont:
-            args.expected_error_rate = 0.15
-            args.ploidy = "haploid"
-            args.ignore_minor_calls = True
-            logger.warning("Setting ploidy to haploid (because --ont flag used)")
-            logger.warning("Setting ignore_minor_calls to True (because --ont flag was used)")
-            logger.warning(f"Setting expected_error_rate error rate to {args.expected_error_rate} (because --ont flag was used)")
-            args.model = "kmer_count"
-
-        # If the user didn't specify the conf_percent_cutoff, then set it
-        # depending on whether or not the --ont flag was used
-        if args.conf_percent_cutoff == -1:
-            if args.ont:
-                args.conf_percent_cutoff = 90
-                logger.warning("Setting conf_percent_cutoff to 90 (was not specified, and --ont flag was used)")
-            else:
-                args.conf_percent_cutoff = 100
-                logger.warning("Setting conf_percent_cutoff to 100 (was not specified, and --ont flag was not used)")
-
+            args.expected_error_rate = ONT_E_RATE
+            logger.info(f"Set expected error rate to {args.expected_error_rate} because --ont flag was used")
+            args.ploidy = ONT_PLOIDY
+            logger.info(f"Set ploidy to {args.ploidy} because --ont flag used")
 
         # conf_percent_cutoff == 100 means that we want to keep all variant calls,
         # in which case there is no need to run the simulations

src/mykrobe/cmds/genotype.py

@@ -7,6 +7,7 @@
 from mykrobe.typing import Genotyper
 from mykrobe.utils import load_json
 from mykrobe.version import __version__
+from mykrobe import ONT_E_RATE
 
 logger = logging.getLogger(__name__)
 
@@ -15,15 +16,10 @@ def run_main(parser, args):
     args = parser.parse_args()
     verbose = True
     if args.ont:
-        args.expected_error_rate = 0.15
-        args.filters = ["LOW_GT_CONF"]
-        args.model = "kmer_count"
+        args.expected_error_rate = ONT_E_RATE
         logger.debug(
             "Setting expected error rate to %s (--ont)" % args.expected_error_rate
         )
-        logger.debug(
-            "Removing LOW_PERCENT_COVERAGE filter (increases sensitivity - in particular for ONT data)"
-        )
 
     if args.min_variant_conf is None:
         args.min_variant_conf = 100

src/mykrobe/constants.py

@@ -3,3 +3,6 @@
 DB_PREFIX = "atlas"
 STOP = "*"
 K = os.environ.get("KMER_SIZE", 21)
+ONT_E_RATE = 0.08
+ONT_PLOIDY = "haploid"
+ILLUMINA_E_RATE = 0.001

src/mykrobe/parser.py

@@ -181,8 +181,12 @@ def __init__(self, *args, **kwargs):
     "--conf_percent_cutoff",
     metavar="conf_percent_cutoff",
     type=float,
-    help="Number between 0 and 100. Determines --min_variant_conf, by simulating variants and choosing the cutoff that would keep x%% of the variants. Default is 90 if --ont, otherwise --min_variant_conf is used as the cutoff",
-    default=-1,
+    help=(
+        "Number between 0 and 100. Determines --min_variant_conf, by simulating "
+        "variants and choosing the cutoff that would keep x%% of the variants "
+        "(default: %(default)d)"
+    ),
+    default=100,
 )
 parser_amr.add_argument(
     "-O",