Set min-length filters by segment (#38)

nextstrain · Oct 3, 2024 · 53a0112 · 53a0112
1 parent 2f644c2
commit 53a0112
Show file tree

Hide file tree

Showing 2 changed files with 4 additions and 2 deletions.
diff --git a/phylogenetic/defaults/config.yaml b/phylogenetic/defaults/config.yaml
@@ -13,7 +13,9 @@ strain_id_field: "accession"
 display_strain_field: "strain"
 
 filter:
-  min_length: 500
+  min_length:
+    l: 5000
+    s: 2000
   include: "defaults/include.txt" # need to always include root strains
   exclude: "defaults/exclude.txt"
   query: "is_lab_host != 'true'"

diff --git a/phylogenetic/rules/prepare_sequences.smk b/phylogenetic/rules/prepare_sequences.smk
@@ -68,7 +68,7 @@ rule filter:
         "benchmarks/{segment}/filter.txt"
     params:
         strain_id_field = config["strain_id_field"],
-        min_length = config['filter']['min_length'],
+        min_length = lambda wildcards: config['filter']['min_length'][wildcards.segment],
         query = config['filter']['query'],
         custom_params = config['filter']['custom_params'],
     shell: