Introduction Usage Output Results

Introduction

ncrnannotator annotates non-coding RNA in genome assemblies using Infernal and the Rfam database. It produces annotation files in GTF, GFF3, and BED format.

Quick start

nextflow run nf-core/ncrnannotator \
  --fasta genome.fa \
  --mode ensembl-vertebrates \
  --rfam_cm Rfam.cm \
  --rfam_seed Rfam.seed \
  --outdir results \
  -profile docker

Required inputs

ParameterDescription
--fastaGenome assembly in FASTA format (uncompressed or .gz)
--modeAnnotation mode (see below)
--rfam_cmRfam covariance model file (Rfam.cm)
--rfam_seedRfam seed alignment file (Rfam.seed)
--outdirOutput directory

Obtaining Rfam files

Download the latest Rfam release from the Rfam FTP:

wget https://ftp.ebi.ac.uk/pub/databases/Rfam/CURRENT/Rfam.cm.gz
wget https://ftp.ebi.ac.uk/pub/databases/Rfam/CURRENT/Rfam.seed.gz
gunzip Rfam.cm.gz Rfam.seed.gz

Annotation modes

ensembl-vertebrates

Annotates vertebrate genomes using a curated subset of Rfam families (snRNA, snoRNA, rRNA, SRP RNA, Y RNA, RNase P, Vault RNA). Uses 1 Mbp genome chunks.

nextflow run nf-core/ncrnannotator \
  --fasta Homo_sapiens.GRCh38.fa \
  --mode ensembl-vertebrates \
  --rfam_cm Rfam.cm \
  --rfam_seed Rfam.seed \
  --outdir results \
  -profile docker

ensembl-invertebrates

Annotates invertebrate genomes using a curated subset of Rfam families. Uses 100 kbp genome chunks for higher sensitivity on compact genomes.

nextflow run nf-core/ncrnannotator \
  --fasta Caenorhabditis_elegans.WBcel235.fa \
  --mode ensembl-invertebrates \
  --rfam_cm Rfam.cm \
  --rfam_seed Rfam.seed \
  --outdir results \
  -profile docker

mgnify-assembly

Annotates metagenomic assemblies using the full Rfam database (no clade filtering). Includes prokaryotic rRNA (bacterial, archaeal). Uses 50 Mbp genome chunks.

nextflow run nf-core/ncrnannotator \
  --fasta metagenome_assembly.fa \
  --mode mgnify-assembly \
  --rfam_cm Rfam.cm \
  --rfam_seed Rfam.seed \
  --outdir results \
  -profile docker

full

Annotates any genome using the complete Rfam database (no clade filtering) with standard Infernal covariance model scoring. Eukaryotic output only (prokaryotic rRNA excluded). Uses 1 Mbp genome chunks.

nextflow run nf-core/ncrnannotator \
  --fasta genome.fa \
  --mode full \
  --rfam_cm Rfam.cm \
  --rfam_seed Rfam.seed \
  --outdir results \
  -profile docker

Optional parameters

ParameterDescriptionDefault
--chunk_sizeOverride genome chunk size (bp)Mode-dependent
--rfam_accessionsCustom Rfam accession list (overrides bundled lists)Bundled per mode

Custom chunk size

For very large genomes or limited memory, override the chunk size:

# Smaller chunks use less memory per cmsearch job
--chunk_size 500000

Running on a laptop

When running locally, use a custom config to cap resource usage:

nextflow run nf-core/ncrnannotator \
  --fasta genome.fa \
  --mode ensembl-invertebrates \
  --rfam_cm Rfam.cm \
  --rfam_seed Rfam.seed \
  --outdir results \
  -c conf/local.config \
  -profile conda

The bundled conf/local.config limits memory to 20 GB and 8 CPUs. Edit it to match your machine.

Using a params file

Rather than specifying all parameters on the command line, you can use a YAML params file:

nextflow run nf-core/ncrnannotator -params-file params.yaml -profile docker

An example params file is provided at assets/params_example.yaml.

Warning

Do not use -c <file> to specify pipeline parameters — only use -c for resource tuning. Use -params-file for parameter overrides.

Resuming a run

Add -resume to restart from the last successful step:

nextflow run nf-core/ncrnannotator \
  --fasta genome.fa \
  --mode ensembl-vertebrates \
  --rfam_cm Rfam.cm \
  --rfam_seed Rfam.seed \
  --outdir results \
  -profile docker \
  -resume

Running on an HPC cluster

SLURM example

nextflow run nf-core/ncrnannotator \
  --fasta genome.fa \
  --mode ensembl-vertebrates \
  --rfam_cm Rfam.cm \
  --rfam_seed Rfam.seed \
  --outdir results \
  -profile singularity \
  -resume

Run Nextflow itself in a SLURM job to avoid timeouts on the head node:

#!/bin/bash
#SBATCH --job-name=ncrnannotator
#SBATCH --time=24:00:00
#SBATCH --mem=8G
#SBATCH --cpus-per-task=2
 
nextflow run nf-core/ncrnannotator \
  -params-file params.yaml \
  -profile singularity \
  -resume

Reproducibility

Pin the pipeline version with -r:

nextflow run nf-core/ncrnannotator -r 1.0.0 \
  --fasta genome.fa \
  --mode ensembl-vertebrates \
  --rfam_cm Rfam.cm \
  --rfam_seed Rfam.seed \
  --outdir results \
  -profile docker

Core Nextflow arguments

-profile

Select a software environment profile:

  • docker — Docker containers (recommended)
  • singularity — Singularity containers (recommended for HPC)
  • conda — Conda environments (fallback when containers unavailable)
  • podman, apptainer, charliecloud — Alternative container runtimes
  • test — Minimal test run (no input files needed)

Example with multiple profiles:

-profile test,docker

-resume

Restart from cached intermediate results. See the Nextflow resume docs.

-c

Provide a custom config for resource tuning only:

-c conf/local.config

Nextflow memory requirements

Limit Nextflow’s own JVM memory usage (add to ~/.bashrc):

export NXF_OPTS='-Xms1g -Xmx4g'
On this page