Consequence Prediction

How vibe-vep predicts variant consequences using Sequence Ontology terms.

Supported Consequences

The tool predicts variant consequences using Sequence Ontology (SO) terms, ordered by impact:

ImpactConsequences
HIGHstop_gained, frameshift_variant, stop_lost, start_lost, splice_donor_variant, splice_acceptor_variant
MODERATEmissense_variant, inframe_insertion, inframe_deletion
LOWsynonymous_variant, splice_region_variant, stop_retained_variant
MODIFIERintron_variant, 5_prime_UTR_variant, 3_prime_UTR_variant, upstream/downstream_gene_variant, non_coding_transcript_exon_variant, mature_miRNA_variant

How It Works

For each variant, the tool determines which transcripts overlap the variant position, then classifies the effect on each transcript:

  1. Upstream/Downstream: Variant outside transcript boundaries
  2. Intronic: Variant between exons. Checks for splice site overlap:
    • Splice donor/acceptor (HIGH): within +/-1-2bp of exon boundary on intron side (donor at 5’ end of intron, acceptor at 3’ end, strand-aware)
    • Splice region (LOW): within 3bp exon side or 3-8bp intron side of splice junction
  3. UTR: Variant in 5’ or 3’ untranslated region
  4. Coding: Variant in CDS — calculates codon/amino acid change to determine missense, synonymous, stop_gained, etc.

For indels, the tool checks the entire deletion span (not just the start position) for:

  • Splice site overlap across the full [pos, pos+len(ref)-1] range
  • Start codon deletion (for deletions spanning from UTR into CDS)
  • Stop codon overlap (frameshift at stop codon produces frameshift_variant,stop_lost)
  • In-frame insertions that create stop codons

Transcript Biotype Handling

The tool treats transcripts as protein-coding if they have defined CDS coordinates (CDSStart > 0 && CDSEnd > 0), which covers:

  • protein_coding
  • nonsense_mediated_decay (appends NMD_transcript_variant modifier)
  • IG_*_gene / TR_*_gene (immunoglobulin/T-cell receptor segments)
  • protein_coding_LoF, non_stop_decay

For miRNA biotype transcripts, exonic variants are classified as mature_miRNA_variant instead of non_coding_transcript_exon_variant.