ViroSync

ViroSync identifies candidate endogenous viral elements (EVEs) in assembled eukaryotic genomes. It searches for viral markers, refines candidate boundaries with gene taxonomy and host-like sequence signals, and scores the resulting regions.

Workflow

ViroSync processes each genome in four phases:

1. Predict proteins after optional repeat masking.
2. Find and validate viral markers, then assemble seed regions.
3. Refine region boundaries with gene taxonomy and host-like sequence signals.
4. Score each candidate and write the accepted EVE set and reports.

Install ViroSync

ViroSync supports Linux x86-64 and uses Pixi for its pinned environment.

git clone https://github.com/NeLLi-team/virosync.git
cd virosync
pixi install --locked
pixi run setup-virosync-resources

Resource setup downloads and verifies resources_v1_0_7_runtime.tar.gz. See Get started for the installation checks.

Quick start

Run the shipped example:

pixi run example

The task writes to results/example/. Check the batch summary:

cat results/example/batch_summary.tsv

The test-1 row must have status=success, predictions=6, and accepted=1. The resource tree records the database version separately:

cat resources/virosync/DB_VERSION

It must print v1.0.7.

Run your own genomes

Give -i a single FASTA file:

pixi run virosync \
  -i genome.fna \
  -o results/my_genome \
  --config config/orchestration.yaml \
  -w 1 \
  --threads-per-worker 16

Or a directory of FASTA files:

pixi run virosync \
  -i genomes/ \
  -o results/my_genomes \
  --config config/orchestration.yaml \
  -w 4 \
  --threads-per-worker 16

ViroSync reads .fna, .fasta, and .fa files. It does not search subdirectories. -i also accepts a text file with one FASTA path per line. -w sets how many genomes run at the same time. --threads-per-worker sets the threads for each of those genomes.

The documentation covers input modes, command-line options, outputs, performance tests, and optional analyses.

ViroSync is available for non-commercial use under the terms in LICENSE.