vep plugins

[BarryDigby]

Description of the bug

Came across this when trying to figure out why a VEP summary html file was missing the summary plots, which I was not able to replicate...

I think that the LoF, SpliceAI and dbNSFP plugins might not be working properly for VEP.

Please find attached a .zip directory containing the workflow logs, and a script I used to re-generate the warning messages. Below is the main body of the .command.sh script for annotating using VEP:

vep \
  --assembly GRCh38 \
  --cache \
  --cache_version 106 \
  --dir_cache /.vep \
  --everything \
  --filter_common \
  --fork 6 \
  --format vcf \
  --input_file HG00138N.haplotypecaller.filtered.vcf.gz \
  --offline \
  --output_file HG00138N.haplotypecaller.filtered_VEP.ann.vcf \
  --per_gene \
  --plugin dbNSFP,ALL,dbNSFP4.2a_grch38.gz,rs_dbSNP,HGVSc_VEP,HGVSp_VEP,1000Gp3_EAS_AF,1000Gp3_AMR_AF,LRT_score,GERP++_RS,gnomAD_exomes_AF \
  --plugin LoF,loftee_path:/opt/conda/envs/nf-core-vep-106.1/share/ensembl-vep-106.1-0 \
  --plugin SpliceAI,snv=spliceai_scores.raw.snv.hg38.vcf.gz,spliceai_scores.raw.indel.hg38.vcf.gz \
  --plugin SpliceRegion \
  --species homo_sapiens \
  --stats_file HG00138N.haplotypecaller.filtered_VEP.summary.html \
  --total_length \
  --vcf

1. dbNSFP consequence

Currently, params.dbnsfp_consequence is pasted without appending consequence= to the parameter value.

--plugin dbNSFP,ALL,dbNSFP4.2a_grch38.gz,rs_dbSNP,HGVSc_VEP,HGVSp_VEP,1000Gp3_EAS_AF,1000Gp3_AMR_AF,LRT_score,GERP++_RS,gnomAD_exomes_AF

Gives the warning:

WARNING: Failed to instantiate plugin dbNSFP: ERROR: Could not retrieve dbNSFP version from filename ALL

i.e the dbNSFP plugin thinks that the consequence string (e.g 'ALL') is the database file.

Solution

I think the lines below:

sarek/conf/modules.config

Line 1211 in 5bb160d

(params.vep_dbnsfp && params.dbnsfp && params.dbnsfp_consequence) ? "--plugin dbNSFP,'${params.dbnsfp_consequence}',${params.dbnsfp.split("/")[-1]},${params.dbnsfp_fields}" : '',

should be:

(params.vep_dbnsfp && params.dbnsfp && params.dbnsfp_consequence)     ? "--plugin dbNSFP,'consequence=${params.dbnsfp_consequence}',${params.dbnsfp.split("/")[-1]},${params.dbnsfp_fields}"               : '',

2. LoF

Nothing to fix here, the LoF plugin will not run because the Bio::perl module is not installed:

WARNING: Failed to compile plugin LoF: Can't locate Bio/Perl.pm in @INC (you may need to install the Bio::Perl module) (@INC contains: /home/bdigby/.vep/Plugins /opt/conda/envs/nf-core-vep-106.1/share/ensembl-vep-106.1-0/modules /opt/conda/envs/nf-core-vep-106.1/share/ensembl-vep-106.1-0 /opt/conda/envs/nf-core-vep-106.1/lib/perl5/5.32/site_perl /opt/conda/envs/nf-core-vep-106.1/lib/perl5/site_perl /opt/conda/envs/nf-core-vep-106.1/lib/perl5/5.32/vendor_perl /opt/conda/envs/nf-core-vep-106.1/lib/perl5/vendor_perl /opt/conda/envs/nf-core-vep-106.1/lib/perl5/5.32/core_perl /opt/conda/envs/nf-core-vep-106.1/lib/perl5/core_perl .) at /opt/conda/envs/nf-core-vep-106.1/share/ensembl-vep-106.1-0/LoF.pm line 46.
BEGIN failed--compilation aborted at /opt/conda/envs/nf-core-vep-106.1/share/ensembl-vep-106.1-0/LoF.pm line 46.
Compilation failed in require at (eval 42) line 2.
BEGIN failed--compilation aborted at (eval 42) line 2.

I don't know enough about LoF vs. Loftool, but the Loftool plugin works (with LoFtool_scores.txt) in the current VEP container provided with the workflow.

3. SpliceAI

--plugin SpliceAI,snv=spliceai_scores.raw.snv.hg38.vcf.gz,spliceai_scores.raw.indel.hg38.vcf.gz 

Gives the warning:

WARNING: Failed to instantiate plugin SpliceAI: ERROR: Failed to parse parameter spliceai_scores.raw.indel.hg38.vcf.gz

Solution

I think it could be as simple as changing:

sarek/conf/modules.config

Line 1213 in 5bb160d

(params.vep_spliceai && params.spliceai_snv && params.spliceai_indel) ? "--plugin SpliceAI,snv=${params.spliceai_snv.split("/")[-1]},${params.spliceai_indel.split("/")[-1]}" : '',

to

(params.vep_spliceai && params.spliceai_snv && params.spliceai_indel) ? "--plugin SpliceAI,snv=${params.spliceai_snv.split("/")[-1]},indel=${params.spliceai_indel.split("/")[-1]}"                  : '',

---

Command used and terminal output

No response

Relevant files

veptest.zip

System information

• nextflow 22.04.5
• AWS
• awsbatch?
• Docker
• ?
• nf-core/sarek 3.0