SnpEff

Annotate VCF variants with predicted functional effects — missense, stop gained, frameshift, splice-site disruption, and more.

What it does

SnpEff maps each variant to the transcripts it overlaps and assigns:

• Effect — missense_variant, stop_gained, splice_donor_variant, etc. (Sequence Ontology terms)
• Impact — HIGH, MODERATE, LOW, or MODIFIER
• Gene / transcript — gene symbol, Ensembl ID, HGVS notation (coding + protein)

Results are written to the ANN INFO field of the output VCF so every downstream tool can read them.

Requirements

Dependency	Why
Java ≥ 8	SnpEff is a Java application
snpEff.jar	The SnpEff JAR — configure or download inside Liatir
Genome database	Per-genome annotation data (snpEffectPredictor.bin)

Setup

Step 1 — Configure the JAR

You can either:

• Browse for an existing snpEff.jar on your machine, or
• Download the latest SnpEff bundle from the official source directly inside Liatir. The download runs in the background with progress, speed, and pause/resume support.

The JAR path is saved by Liatir and persists across restarts.

Step 2 — Download a genome database

Select a genome from the dropdown or type a custom ID (e.g. GRCh38.p14). Liatir runs:

java -jar snpEff.jar download <genome> -dataDir <dir>

Available built-in genomes:

ID	Build
hg38	Human GRCh38
hg19	Human GRCh37
GRCh38.105	Human GRCh38.105 (Ensembl)
mm39	Mouse GRCm39
mm10	Mouse GRCm38
rn7	Rat mRatBN7.2
danRer11	Zebrafish GRCz11
dm6	Drosophila BDGP6
ce11	C. elegans WBcel235
sacCer3	Yeast R64

Running annotation

With JAR and genome ready, select a VCF/VCF.gz file and click Annotate. Liatir runs:

java -Xmx4g -jar snpEff.jar ann \
  -dataDir <dir> \
  -noStats -noLog \
  <genome> \
  <input.vcf>

The annotated VCF appears in the results panel and can be added to the Data library.

Output

Summary stats

Stat	Description
Total variants	All records processed
HIGH impact	Stop gained, frameshift, splice site
MODERATE impact	Missense, in-frame indel
LOW impact	Synonymous, splice region

ANN field

Each variant gets an ANN= INFO field with one entry per overlapping transcript:

ANN=A|missense_variant|MODERATE|BRCA1|ENSG00000012048|
    transcript|ENST00000357654.9|protein_coding|
    18/23|c.5266dupC|p.Gln1756fs|...

Pipe-separated fields (simplified): allele | effect | impact | gene name | gene ID | feature type | feature ID | biotype | exon rank | HGVS.c | HGVS.p | …

Pipeline use

{
  "id": "annotate-variants",
  "type": "native-tool",
  "tool": "snpeff",
  "inputs": {
    "inputFile": "{{ outputs.filter.vcf }}",
    "genome": "hg38"
  }
}

Troubleshooting

Java not found — install Java 8+ from your package manager and make sure java is on PATH.

Database download fails — SnpEff fetches databases from snpeff.sourceforge.net. Check your network or firewall settings.

Out of memory — the default heap is -Xmx4g. For very large VCFs on memory-constrained machines consider closing other apps first.

Wrong genome ID — SnpEff genome IDs are case-sensitive. Check the full list at snpeff.sourceforge.net/download.html.