builds.yaml

custom_rules:
  - nextstrain_profiles/nextstrain/subsampling_ranges.smk
  - ecdc_profile/localrules.smk

inputs:
  - name: gisaid
    metadata: data/GISAID/metadata.tsv
    sequences: data/GISAID/sequences.fasta


builds:
  # with a build name that will produce the following URL fragment on Nextstrain/auspice:
  # /ncov/Europe/Sweden
  sweden:
      subsampling_scheme: custom_country
      geographic_scale: country
      region: Europe
      country: Sweden
      subclades: "nextstrain_profiles/nextstrain/subclades_EU.tsv"

# Here, you can specify what type of auspice_config you want to use
# and what description you want. These will apply to all the above builds.
# If you want to specify specific files for each build - you can!
# See the 'example_advanced_customization' builds.yaml
files:
  colors: "ecdc_profile/colors.tsv"
  auspice_config: "ecdc_profile/auspice_config_ecdc.json"
  description: "ecdc_profile/description.md"

subsampling:
# Default subsampling logic for countries
  custom_country:
    # Focal samples for country
    country_early:
      group_by: "division year month"
      max_sequences: 500
      exclude: "--exclude-where 'country!={country}'"
    country_late:
      group_by: "division year month"
      max_sequences: 1000
      exclude: "--exclude-where 'country!={country}'"
    # Contextual samples from country's region
    region_early:
      group_by: "country year month"
      max_sequences: 250
      exclude: "--exclude-where 'country={country}' 'region!={region}'"
      priorities:
        type: "proximity"
        focus: "country_early"
    # Contextual samples from the rest of the world,
    # excluding the current region to avoid resampling.
    global_early:
      group_by: "country year month"
      max_sequences: 300
      exclude: "--exclude-where 'region={region}'"
      priorities:
        type: "proximity"
        focus: "country_early"
    region_late:
      group_by: "country year month"
      max_sequences: 750
      exclude: "--exclude-where 'country={country}' 'region!={region}'"
      priorities:
        type: "proximity"
        focus: "country_late"
    # Contextual samples from the rest of the world,
    # excluding the current region to avoid resampling.
    global_late:
      group_by: "country year month"
      max_sequences: 500
      exclude: "--exclude-where 'region={region}'"
      priorities:
        type: "proximity"
        focus: "country_late"