I'm looking for COSMIC v98 vcfs/tsv files for GRCh37. This is because the current version of Variant Effect Predictor (VEP v115 and v116) use COSMIC v98 for GRCh37. However, I cannot find them on https://cancer.sanger.ac.uk, https://genome.ucsc.edu, or on the Wayback machine. This is because only the newer versions are listed for COSMIC (v101 or later). If you have a better idea to solve my problem, feel free to post that too. Thank you.
1 answer
I reached out to COSMIC/Qiagen and VEP. Silence from Qiagen. VEP at least opened a ticket for me, but I haven't heard back in a week. After doing some reading:
- Retrieving allele-specific information for a variant using VEP annotation
- https://jun2026.archive.ensembl.org/info/docs/tools/vep/script/vep_custom.html
I decided that the default COSMIC annotations from VEP were not correct anyway because they're not allele specific (they are loci specific due to licensing limitations) so I'm trying to do allele specific COSMIC annotations. I couldn't find the old v98 COSMIC annotations so I went with v104 (latest version). There's some extra stuff in the code diff: (A) --pick_order rank so I can get the best effect (B) a collaborator requested COSMIC 80 counts. This isn't the cleanest solution because the COSV annotations are still repeated for every comma separated CSQ (consequence) output by VEP. Also, I have yet to remove the v98 COSMIC annotations (I think I'll do that when I generate an excel report). For those of you running the Sarek Pipeline, this is the diff:
diff --git a/conf/ccg/ccg.config b/conf/ccg/ccg.config
index 711af673c..7159aec39 100644
--- a/conf/ccg/ccg.config
+++ b/conf/ccg/ccg.config
@@ -11,7 +11,20 @@ params {
// "--flag_pick" = Adds PICK=1 to picked annotation block
// "--pick_order rank" = Order by most deleterious effect (i.e. best effect)
// The others were already there in Sarek, but disappeared when I added the vep_custom_args
- vep_custom_args = "--flag_pick --pick_order rank --everything --filter_common --offline --total_length"
+ vep_custom_args = [
+ "--flag_pick --pick_order rank --everything --filter_common --offline --total_length",
+ "--custom Cosmic_GenomeScreensMutant_v104_GRCh37.vcf.gz,COSMIC_v104,vcf,exact,0,ID,GENOME_SCREEN_SAMPLE_COUNT",
+ "--custom Cosmic_CompleteTargetedScreensMutant_v104_GRCh37.vcf.gz,COSMIC_v104,vcf,exact,0,TARGETED_SCREEN_SAMPLE_COUNT",
+ "--custom CosmicCodingMuts_v80_normalised_v1.sorted.SNPremoved.vcf.gz,COSMIC_v80,vcf,exact,0,ID,CNT"
+ ].join(" ")
+ vep_extra_files = [
+ "gs://ccg-out/mnt/storage/centers/ccgd/reference_files/genomes/homo_sapiens/GRCh37/cosmic/Cosmic_CompleteTargetedScreensMutant_v104_GRCh37.vcf.gz",
+ "gs://ccg-out/mnt/storage/centers/ccgd/reference_files/genomes/homo_sapiens/GRCh37/cosmic/Cosmic_CompleteTargetedScreensMutant_v104_GRCh37.vcf.gz.tbi",
+ "gs://ccg-out/mnt/storage/centers/ccgd/reference_files/genomes/homo_sapiens/GRCh37/cosmic/Cosmic_GenomeScreensMutant_v104_GRCh37.vcf.gz",
+ "gs://ccg-out/mnt/storage/centers/ccgd/reference_files/genomes/homo_sapiens/GRCh37/cosmic/Cosmic_GenomeScreensMutant_v104_GRCh37.vcf.gz.tbi",
+ "gs://ccg-out/mnt/storage/centers/ccgd/reference_files/genomes/homo_sapiens/GRCh37/cosmic/CosmicCodingMuts_v80_normalised_v1.sorted.SNPremoved.vcf.gz",
+ "gs://ccg-out/mnt/storage/centers/ccgd/reference_files/genomes/homo_sapiens/GRCh37/cosmic/CosmicCodingMuts_v80_normalised_v1.sorted.SNPremoved.vcf.gz.tbi"
+ ].join(",")
}
profiles {
diff --git a/main.nf b/main.nf
index 7c5a4ddd2..8d0daf294 100755
--- a/main.nf
+++ b/main.nf
@@ -223,6 +223,12 @@ workflow NFCORE_SAREK {
vep_extra_files.add(file(params.spliceai_snv_tbi, checkIfExists: true))
}
+ if (params.vep_extra_files) {
+ params.vep_extra_files.split(",").each { f ->
+ vep_extra_files.add(file(f, checkIfExists: true))
+ }
+ }
+
//
// WORKFLOW: Run pipeline
//
Log in to answer this question.
Unrelated, why are you still using GRCh37? It's prehistoric at this point.
We use both. It's just for older, ongoing projects where new data needs to be compared to old data.
Have you tried writing to support at COSMIC to see if they can make the download available? If you are from a commercial entity then you will need to contact Qiagen.