feat: upgrade dbNSFP to v4.5 (#77) (#78)

varfish-org · Jan 1, 2024 · 62d2f9a · 62d2f9a
1 parent 53e4209
commit 62d2f9a
Show file tree

Hide file tree

Showing 10 changed files with 171 additions and 27 deletions.
diff --git a/download_urls.yml b/download_urls.yml
@@ -1,3 +1,19 @@
+# dbNSFP v4.5a
+- url: https://dbnsfp.s3.amazonaws.com/dbNSFP4.5a.zip
+  excerpt_strategy:
+    strategy: manual
+    count: null
+# dbNSFP v4.5c
+- url: https://dbnsfp.s3.amazonaws.com/dbNSFP4.5c.zip
+  excerpt_strategy:
+    strategy: manual
+    count: null
+- url: ftp://dbnsfp:[email protected]/dbscSNV1.1.zip
+  skip_upstream_check: true  # does not work reliably in tests
+  excerpt_strategy:
+    strategy: manual
+    count: null
+
 - url: https://storage.googleapis.com/dm_alphamissense/AlphaMissense_hg19.tsv.gz
 - url: https://storage.googleapis.com/dm_alphamissense/AlphaMissense_hg38.tsv.gz
 - url: https://storage.googleapis.com/dm_alphamissense/AlphaMissense_gene_hg38.tsv.gz
@@ -168,20 +184,6 @@
 - url: https://kircherlab.bihealth.org/download/CADD/v1.6/GRCh38/gnomad.genomes.r3.0.indel_inclAnno.tsv.gz
 - url: https://kircherlab.bihealth.org/download/CADD/v1.6/GRCh38/gnomad.genomes.r3.0.indel_inclAnno.tsv.gz.tbi
 
-- url: https://usf.box.com/shared/static/bvfzmkpgtphvbmmrvb2iyl2jl21o49kc
-  excerpt_strategy:
-    strategy: manual
-    count: null
-- url: https://usf.box.com/shared/static/a84zcdlkx2asq2nxh6xr2gdb4csmyvhk
-  excerpt_strategy:
-    strategy: manual
-    count: null
-- url: ftp://dbnsfp:[email protected]/dbscSNV1.1.zip
-  skip_upstream_check: true  # does not work reliably in tests
-  excerpt_strategy:
-    strategy: manual
-    count: null
-
 - url: https://ftp.ncbi.nlm.nih.gov/genomes/archive/old_refseq/H_sapiens/ARCHIVE/ANNOTATION_RELEASE.105/Assembled_chromosomes/chr_accessions_GRCh37.p13
   excerpt_strategy:
     strategy: no-excerpt

diff --git a/excerpt-data/8bad0386c4cd562f/dbNSFP4.5c.zip b/excerpt-data/8bad0386c4cd562f/dbNSFP4.5c.zip
diff --git a/excerpt-data/8bad0386c4cd562f/url.txt b/excerpt-data/8bad0386c4cd562f/url.txt
diff --git a/excerpt-data/b920970c3e3329fb/dbNSFP4.5a.zip b/excerpt-data/b920970c3e3329fb/dbNSFP4.5a.zip
diff --git a/excerpt-data/b920970c3e3329fb/url.txt b/excerpt-data/b920970c3e3329fb/url.txt
diff --git a/...s/output/annonars/dbnsfp-schema-4.4a.json → ...s/output/annonars/dbnsfp-schema-4.5a.json b/...s/output/annonars/dbnsfp-schema-4.4a.json → ...s/output/annonars/dbnsfp-schema-4.5a.json
@@ -484,6 +484,82 @@
       "name": "VARITY_ER_LOO_rankscore",
       "typ": "Float"
     },
+    {
+      "name": "ESM1b_score",
+      "typ": "String"
+    },
+    {
+      "name": "ESM1b_rankscore",
+      "typ": "Float"
+    },
+    {
+      "name": "ESM1b_pred",
+      "typ": "String"
+    },
+    {
+      "name": "EVE_score",
+      "typ": "String"
+    },
+    {
+      "name": "EVE_rankscore",
+      "typ": "Float"
+    },
+    {
+      "name": "EVE_Class10_pred",
+      "typ": "String"
+    },
+    {
+      "name": "EVE_Class20_pred",
+      "typ": "String"
+    },
+    {
+      "name": "EVE_Class25_pred",
+      "typ": "String"
+    },
+    {
+      "name": "EVE_Class30_pred",
+      "typ": "String"
+    },
+    {
+      "name": "EVE_Class40_pred",
+      "typ": "String"
+    },
+    {
+      "name": "EVE_Class50_pred",
+      "typ": "String"
+    },
+    {
+      "name": "EVE_Class60_pred",
+      "typ": "String"
+    },
+    {
+      "name": "EVE_Class70_pred",
+      "typ": "String"
+    },
+    {
+      "name": "EVE_Class75_pred",
+      "typ": "String"
+    },
+    {
+      "name": "EVE_Class80_pred",
+      "typ": "String"
+    },
+    {
+      "name": "EVE_Class90_pred",
+      "typ": "String"
+    },
+    {
+      "name": "AlphaMissense_score",
+      "typ": "String"
+    },
+    {
+      "name": "AlphaMissense_rankscore",
+      "typ": "Float"
+    },
+    {
+      "name": "AlphaMissense_pred",
+      "typ": "String"
+    },
     {
       "name": "Aloft_Fraction_transcripts_affected",
       "typ": "String"

diff --git a/...s/output/annonars/dbnsfp-schema-4.4c.json → ...s/output/annonars/dbnsfp-schema-4.5c.json b/...s/output/annonars/dbnsfp-schema-4.4c.json → ...s/output/annonars/dbnsfp-schema-4.5c.json
@@ -432,6 +432,70 @@
       "name": "VARITY_ER_LOO_rankscore",
       "typ": "Float"
     },
+    {
+      "name": "ESM1b_score",
+      "typ": "String"
+    },
+    {
+      "name": "ESM1b_rankscore",
+      "typ": "Float"
+    },
+    {
+      "name": "ESM1b_pred",
+      "typ": "String"
+    },
+    {
+      "name": "EVE_score",
+      "typ": "String"
+    },
+    {
+      "name": "EVE_rankscore",
+      "typ": "Float"
+    },
+    {
+      "name": "EVE_Class10_pred",
+      "typ": "String"
+    },
+    {
+      "name": "EVE_Class20_pred",
+      "typ": "String"
+    },
+    {
+      "name": "EVE_Class25_pred",
+      "typ": "String"
+    },
+    {
+      "name": "EVE_Class30_pred",
+      "typ": "String"
+    },
+    {
+      "name": "EVE_Class40_pred",
+      "typ": "String"
+    },
+    {
+      "name": "EVE_Class50_pred",
+      "typ": "String"
+    },
+    {
+      "name": "EVE_Class60_pred",
+      "typ": "String"
+    },
+    {
+      "name": "EVE_Class70_pred",
+      "typ": "String"
+    },
+    {
+      "name": "EVE_Class75_pred",
+      "typ": "String"
+    },
+    {
+      "name": "EVE_Class80_pred",
+      "typ": "String"
+    },
+    {
+      "name": "EVE_Class90_pred",
+      "typ": "String"
+    },
     {
       "name": "Aloft_Fraction_transcripts_affected",
       "typ": "String"

diff --git a/rules/output/annonars/dbnsfp.smk b/rules/output/annonars/dbnsfp.smk
@@ -32,9 +32,9 @@ rule output_annonars_dbnsfp:  # -- build dbNSFP RocksDB with annonars
             --genome-release {wildcards.genome_release} \
             --null-values=. \
             --inference-row-count 100000 \
-            --path-schema-json rules/output/annonars/dbnsfp-schema-{wildcards.v_dbnsfp}.json \
             \
             --path-out-rocksdb $(dirname {output.rocksdb_identity}) \
+            --path-schema-json rules/output/annonars/dbnsfp-schema-{wildcards.v_dbnsfp}.json \
             \
             $(if [[ "{wildcards.genome_release}" == "grch37" ]]; then \
                 echo --col-chrom 'hg19_chr'; \

diff --git a/rules/work/annos/seqvars/dbnsfp.smk b/rules/work/annos/seqvars/dbnsfp.smk
@@ -1,12 +1,6 @@
 ## Rules related to dbNSFP.
 
 
-#: Download URL for dbNSFP 4.4a
-DBNSFP_ACADEMIC_URL = "https://usf.box.com/shared/static/bvfzmkpgtphvbmmrvb2iyl2jl21o49kc"
-#: Download URL for dbNSFP 4.4c
-DBNSFP_COMMMERCIAL_URL = "https://usf.box.com/shared/static/a84zcdlkx2asq2nxh6xr2gdb4csmyvhk"
-
-
 def files_dbnsfp():
     """Helper that returns the files within the dbNSFP archive."""
     lst = [
@@ -60,11 +54,7 @@ rule annos_seqvars_dbnsfp_download:  # -- download dbNSFP ZIP file
     threads: 8
     shell:
         r"""
-        if [[ "{wildcards.variant}" == a ]]; then
-            url={DBNSFP_ACADEMIC_URL}
-        else
-            url={DBNSFP_COMMMERCIAL_URL}
-        fi
+        url=https://dbnsfp.s3.amazonaws.com/dbNSFP4.5{wildcards.variant}.zip
 
         aria2c \
             --check-certificate=false \

diff --git a/varfish_db_downloader/versions.py b/varfish_db_downloader/versions.py
@@ -130,7 +130,7 @@ class DataVersions:
     ensembl_38="109",
     ensembl="110",
     today=TODAY,
-    dbnsfp="4.4",
+    dbnsfp="4.5",
     dbscsnv="1.1",
     cadd="1.6",
     gnomad_constraints="4.0",