Refactor annotations out of config, into rule [#8]

ingest/config/defaults.yaml

@@ -11,31 +11,29 @@ viruses:
 # for each virus, give the NCBI taxon ID (required for data fetch from
 # NCBI datasets) and the path to the manual annotations file, relative
 # to the ingest directory
-## TODO see if the `annotations` key can be refactored out into the
-## Snakemake file with a `{virus}` wildcard
 229e:
   ncbi_taxon_id: "11137"
-  annotations: "config/229e/annotations.tsv"
 nl63:
   ncbi_taxon_id: "277944"
-  annotations: "config/nl63/annotations.tsv"
 oc43:
   ncbi_taxon_id: "31631"
-  annotations: "config/oc43/annotations.tsv"
 hku1:
   ncbi_taxon_id: "290028"
-  annotations: "config/hku1/annotations.tsv"
+
 # Optional fields to add to the NCBI Datasets output
 ncbi_dataset_fields: []
+
 # Config parameters related to the curate pipeline, shared by all viruses
 curate:
   # URL pointed to public generalized geolocation rules
   # For the Nextstrain team, this is currently
   # 'https://raw.githubusercontent.com/nextstrain/ncov-ingest/@/source-data/gisaid_geoLocationRules.tsv'
   geolocation_rules_url: "https://raw.githubusercontent.com/nextstrain/ncov-ingest/@/source-data/gisaid_geoLocationRules.tsv"
+
   # The path to the local geolocation rules within the pathogen repo
   # The path should be relative to the ingest directory.
   local_geolocation_rules: "config/geolocation_rules.tsv"
+
   # List of field names to change in the format of <old_field_name>=<new_field_name>
   # This is the first step in the pipeline, so any references to field names
   # in the configs below should use the new field names
@@ -51,8 +49,10 @@ curate:
     Submitter Affiliation: institution
     SRA Accessions: sra_accession
     passage_type: passage_type
+
   # List of date fields to standardize to ISO format YYYY-MM-DD
   date_fields: ["date"]
+
   # List of expected date formats that are present in the date fields provided above
   # These date formats should use directives expected by datetime
   # See https://docs.python.org/3.9/library/datetime.html#strftime-and-strptime-format-codes
@@ -61,6 +61,7 @@ curate:
     - "%Y-%m"
     - "%Y-%m-%d"
     - "%Y-%m-%dT%H:%M:%SZ"
+
   titlecase:
     # Abbreviations not cast to titlecase, keeps uppercase
     abbreviations: ["USA"]
@@ -85,20 +86,28 @@ curate:
       - sur
       - the
       - y
+
     # List of string fields to titlecase
     fields: ["region", "country", "division", "location"]
+
   # Metadata field that contains the list of authors associated with the sequence
   authors_field: "authors"
+
   # Default value to use if the authors field is empty
   authors_default_value: "?"
+
   # Name to use for the generated abbreviated authors field
   abbr_authors_field: "abbr_authors"
+
   # The ID field in the metadata to use to merge the manual annotations
   annotations_id: "strain"
+
   # The ID field in the metadata to use as the sequence id in the output FASTA file
   output_id_field: "strain"
+
   # The field in the NDJSON record that contains the actual genomic sequence
   output_sequence_field: "sequence"
+
   # The list of metadata columns to keep in the final output of the curation pipeline.
   metadata_columns:
     - strain

ingest/rules/curate.smk

@@ -65,7 +65,7 @@ rule curate:
         sequences_ndjson="data/{virus}/ncbi.ndjson",
         # Change the geolocation_rules input path if you are removing the above two rules
         all_geolocation_rules="data/all-geolocation-rules.tsv",
-        annotations=lambda wildcards: config[wildcards.virus]["annotations"],
+        annotations="config/{virus}/annotations.tsv",
     output:
         metadata="results/{virus}/all_metadata.tsv",
         sequences="results/{virus}/sequences.fasta",