Extend EVE plugin to handle popEVE data

[ainefairbrother]

Description

This PR extends the VEP plugin EVE.pm to be able to handle popEVE data. As such, the plugin can now annotate using EVE data OR popEVE data OR both.

Main changes

• Added a new parameter: popeve_file
  • Backwards compatibility enabled by leaving the file parameter as is - this param is still for the EVE file
• Plugin now outputs popEVE_* fields when a popEVE record matches the user's input, and continues to emit EVE_* fields when an EVE record matches
• Added handling of the popEVE file and handling of the matching results due to popEVE/EVE differences:
  • popEVE = genomic SNV scores, so POS is the variant position
  • EVE = codon-level substitutions (3-bp alleles), so POS is the codon start and the user's variant might be at POS, POS+1, POS+2
  • As such, the EVE logic is retained:
    • A codon-sized window 2nt upstream of the input variant is defined, with tabix returning all records present in that range (typically 0|1 popEVE rows and 0-3 EVE rows (1 per possible codon start))
    • Every returned row is tested agains the user's input allele with get_matched_variant_alleles
    • EVE rows match only if the user's change can produce the ALT codon from the REF codon (a 1bp SNV won't match codons requiring 2-3 base changes)
  • New logic added for popEVE and EVE/popEVE merging
    • For popEVE, rows match on the exact SNV at POS
    • All matches found for popEVE & EVE are then merged into a single output row

Testing

• It would be beneficial if the reviewer would test with a range of input variants
• See header of EVE.pm for EVE/popEVE input file download and prep instructions, however, the EVE file is available on nfs at ensembl/variation/data/EVE

in.vcf:

##fileformat=VCFv4.2
##source=test
#CHROM	POS	ID	REF	ALT	QUAL	FILTER	INFO
chr1	961387	.	G	C	.	PASS	.

VEP command:

EVE_file="eve_merged.vcf.gz"
POPEVE_file="grch38_popEVE_ukbb_20250715.vcf.gz"

./vep \
-i in.vcf \
-o out.tsv \
--dir_cache /nfs/production/flicek/ensembl/variation/data/VEP/tabixconverted \
--cache_version 115 \
--format vcf --symbol --tab --offline \
--cache \
--fasta Homo_sapiens.GRCh38.dna.toplevel.fa.gz \
--assembly GRCh38 \
--force_overwrite \
--no_escape \
--show_ref_allele \
--no_stats \
--plugin EVE,file=$EVE_file,popeve_file=$POPEVE_file

Output:

• 76-row TSV file with 2 EVE_ columns and 9 popEVE_ columns
• 7 rows should be annotated with both EVE and popEVE values

[nakib103]

if only popEVE file is given they will be still be in the header.

[ainefairbrother]

I have now hard-coded them and added logic so that only the relevant descriptions (pop, eve or both) end up in the output file.

[nakib103]

it would be better to keep the formatting same as before, as other plugins use the same format.

[ainefairbrother]

I have reverted the styling.

[nakib103]

@ainefairbrother ,

Besides the other comments, it is also better to add -

• an example command for popEVE in the synopsis.
• instead of using Bio::DB::HTS can we hardcode the header? in some cases the user might not have installed the library and it would restrict them to use the plugin.

[ainefairbrother]

@ainefairbrother ,

Besides the other comments, it is also better to add -

• an example command for popEVE in the synopsis.

Done.

• instead of using Bio::DB::HTS can we hardcode the header? in some cases the user might not have installed the library and it would restrict them to use the plugin.

Done.

[nakib103]

Looks good @ainefairbrother
couple of checks to be added

[nakib103]

LGTM !~