BAsic Rapid Ribosomal RNA Predictor
Barrnap predicts the location of ribosomal RNA genes in genomes. It supports bacteria (5S,23S,16S), archaea (5S,5.8S,23S,16S), metazoan mitochondria (12S,16S) and eukaryotes (5S,5.8S,28S,18S).
It takes FASTA DNA sequence as input, and write GFF3 as output.
It uses the new nhmmer tool that comes with HMMER 3.1 for HMM searching in RNA:DNA style.
Multithreading is supported and one can expect roughly linear speed-ups with more CPUs.
- Perl 5.xx (core modules only)
- nhmmer (part of HMMER 3.x)
- bedtools >= 2.27.0
conda install -c bioconda -c conda-forge barrnap
Install Homebrew (macOS) or Linuxbrew (Linux).
brew install brewsci/bio/barrnap
This will install the latest version direct from Github.
You'll need to add the bin directory to your PATH.
cd $HOME
git clone https://github.com/tseemann/barrnap.git
cd barrnap/bin
./barrnap --help
% barrnap --quiet examples/small.fna
##gff-version 3
P.marinus barrnap:0.8 rRNA 353314 354793 0 + . Name=16S_rRNA;product=16S ribosomal RNA
P.marinus barrnap:0.8 rRNA 355464 358334 0 + . Name=23S_rRNA;product=23S ribosomal RNA
P.marinus barrnap:0.8 rRNA 358433 358536 7.5e-07 + . Name=5S_rRNA;product=5S ribosomal RNA
% barrnap -q -k mito examples/mitochondria.fna
##gff-version 3
AF346967.1 barrnap:0.8 rRNA 643 1610 . + . Name=12S_rRNA;product=12S ribosomal RNA
AF346967.1 barrnap:0.8 rRNA 1672 3228 . + . Name=16S_rRNA;product=16S ribosomal RNA
% barrnap -o rrna.fa < contigs.fa > rrna.gff
% head -n 3 rrna.fa
>16S_rRNA::gi|329138943|tpg|BK006945.2|:455935-456864(-)
ACGGTCGGGGGCATCAGTATTCAATTGTCAGAGGTGAAATTCTTGGATT
TATTGAAGACTAACTACTGCGAAAGCATTTGCCAAGGACGTTTTCATTA
--helpshow help and exit--versionprint version in formbarrnap X.Yand exit--citationprint a citation and exit
--kingdomis the database to use: Bacteria:bac, Archaea:arc, Eukaryota:euk, Metazoan Mitochondria:mito--threadsis how many CPUs to assign tonhmmersearch--evalueis the cut-off fornhmmerreporting, before further scrutiny--lencutoffis the proportion of the full length that qualifies aspartialmatch--rejectwill not include hits below this proportion of the expected length
--quietwill not print any messages tostderr--incseqwill include the full input sequences in the output GFF--outseqcreates a FASTA file with the hit sequences
Barrnap does not do anything fancy. It has HMM models for each different rRNA gene. They are built from full length seed alignments.
Barrnap is designed to be a substitute for RNAmmer. It was motivated by my desire to remove Prokka's dependency on RNAmmer which is encumbered by a free-for-academic sign-up license, and by RNAmmer's dependence on legacy HMMER 2.x which conflicts with HMMER 3.x that most people are using now.
RNAmmer is more sophisticated than Barrnap, and more accurate because it uses HMMER 2.x in glocal alignment mode whereas NHMMER 3.x currently only supports local alignment (Sean Eddy expected glocal to be supported in 2014, but it still isn't available in 2018).
In practice, Barrnap will find all the typical rRNA genes in a few seconds (in bacteria), but may get the end points out by a few bases and will probably miss wierd rRNAs. The HMM models it uses are derived from Rfam, Silva and RefSeq.
Bacteria (70S)
LSU 50S
5S RF00001
23S SILVA-LSU-Bac
SSU 30S
16S RF00177
Archaea (70S)
LSU 50S
5S RF00001
5.8S RF00002
23S SILVA-LSU-Arc
SSU 30S
16S RF01959
Eukarya (80S)
LSU 60S
5S RF00001
5.8S RF00002
28S SILVA-LSU-Euk
SSU 40S
18S RF01960
Metazoan Mito
12S RefSeq (MT-RNR1, s-rRNA, rns)
16S RefSeq (MT-RNR2, l-rRNA, rnl)
Fungi [Sajeet Haridas]
LSU 35S ?
5S
5.8S
25S
SSU ?
18S
Mito [http://www.ncbi.nlm.nih.gov/nuccore/NC_001224.1]
15S
21S (multiple exons)
Apicoplast [http://www.ncbi.nlm.nih.gov/nuccore/U87145.2]
LSU ~2500bp 28S ?
SSU ~1500bp 16S ?
Plant [Shaun Jackman]
Mito [https://www.ncbi.nlm.nih.gov/nucleotide?cmd=Retrieve&dopt=GenBank&list_uids=26556996]
5S ~118 bp ? rrn5 (use RF00001 ?)
18S ~1935 bp ? rrn18 (use RF01960 ?)
26S ~2568 bp ? rrn26
The name Barrnap was originally derived from Bacterial/Archaeal Ribosomal RNA Predictor. However it has since been extended to support mitochondrial and eukaryotic rRNAs, and has been given the new backronym BAsic Rapid Ribosomal RNA Predictor. The project was originally spawned at CodeFest 2013 in Berlin, Germany by Torsten Seemann and Tim Booth.
- Barrnap: GPLv3
- Rfam: CC0
- SILVA: Free for academic use
Torsten Seemann
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent