BAsic Rapid Ribosomal RNA Predictor
Torsten Seemann - [email protected] - @torstenseemann
Barrnap predicts the location of ribosomal RNA genes in genomes. It supports bacteria (5S,23S,16S), archaea (5S,5.8S,23S,16S), 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. NHMMER binaries for 64-bit Linux and Mac OS X are included and will be auto-detected. Multithreading is supported and one can expect roughly linear speed-ups with more CPUs.
% cd $HOME
% tar zxvf barrnap-0.X.tar.gz
% echo "PATH=$PATH:$HOME/barrnap-0.x/bin" >> .bashrc
(logout and log back in)
% barrnap --quiet examples/small.fna
##gff-version 3
P.marinus barrnap:0.7 rRNA 353314 354793 0 + . Name=16S_rRNA;product=16S ribosomal RNA
P.marinus barrnap:0.7 rRNA 355464 358334 0 + . Name=23S_rRNA;product=23S ribosomal RNA
P.marinus barrnap:0.7 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:0.7 rRNA 643 1610 . + . Name=12S_rRNA;product=12S ribosomal RNA
AF346967.1 barrnap:0.7 rRNA 1672 3228 . + . Name=16S_rRNA;product=16S ribosomal RNA
Barrnap does not do anything fancy. It has HMM models for each different rRNA gene. They are built from full length seed alignments.
- Perl >= 5.6
- HMMER >= 3.1b
Barrnap is free software, released under the GPL (version 3).
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 expects glocal to be supported in 2014). 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 Mito 12S RefSeq (MT-RNR1, s-rRNA, rns) 16S RefSeq (MT-RNR2, l-rRNA, rnl) TODO: [Sajeet Haridas] Fungi LSU 35S ? 5S 5.8S 25S SSU ? 18S Mito [http://www.ncbi.nlm.nih.gov/nuccore/NC_001224.1] 15S 21S (multiple exons) TODO: Apicoplast [http://www.ncbi.nlm.nih.gov/nuccore/U87145.2] LSU ~2500bp 28S ? SSU ~1500bp 16S ? Plastid [Shaun Jackman] ?
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.