Troubleshooting barrnap

,
1

Why is there a discrepancy between the sequences extracted using Barrnap and the actual biological positions of the rRNA?

2

Welcome @sasdfas

Hopefully we can help!

To clarify, are you asking about why experimental mapping results may be different from what someone else’s results were, or what a published “result” is?

This is a complicated question but in short: the bases in the query are processed through the algorithm to identify predicted regions on that query. This is similar to mapping reads – a query’s bases are compared to the bases in the target, and according to the algorithm’s parameters, a “match” between those are identified. The “match” for this tool is a region on the query that the extra processing compares to a reference that has also gone through some processing (in this case, HMM domains), and this is what produces the “matches”. The matches are then written to an annotation file (GFF) with regions defined on the query and what was found there.

These and other bioinformatic comparisons are always predictions!

Choosing the right tool is important, as is preparing the inputs and setting parameters. Reviewing more about how an algorithm is used for different use cases is usually a good place to start!

The Help section on tool forms will usually have some quick help then link outs to more. The development repository, publications and author notes, and discussions or tutorials available online will contain the details and possibly examples where other people shared what they did.

Help

barrnap

barrnap predicts the location of 5S, 16S and 23S ribosomal RNA genes in Bacterial genome sequences. Barrnap now supports Archaea, Eukaryota and Mitochondria. It takes FASTA DNA sequence as input, and write GFF3 as output. It uses the new NHMMER tool that comes with HMMER 3.1-dev for HMM searching in DNA: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. This tool 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 an free-for-academic sign-up license, and by the needed 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 HMMER 3.x currently only supports local alignment (Sean Eddy expects glocal to be supported in 2014). In practice, Barrnap will find all the typical 5/16/23S genes 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 GreenGenes.

Scroll down to find descriptions of the parameters …

Points to → GitHub - tseemann/barrnap: 🔬 Bacterial ribosomal RNA predictor · GitHub. Tools work the same in Galaxy as they do other places. This means resources outside of Galaxy can be helpful for scientific usage and data interpretation. The author listed out similar tools near the end of the README, so those might be able to provide more context.

If you are new to this type of annotation pipeline, you could explore Galaxy tutorials that do something similar using different tools. These are usually based on a scientific publication, and that is discussed. When a tool is included in tutorials, find those linked from the bottom of the tools form. Example:

You can also search the GTN directly! Or explore Topic categories, and the Learning Pathways.

Finally, to locate these similar tools and resources, you can also try an Advanced Search in the tool panel!

  • Activity bar → Tools → Discover Tools

There are many resources and I hope this gets you started! I wasn’t sure if you had an error or just an unexpected result. We can help to troubleshooting technical issues if that is happening, or if you are not sure, please let us know! We would probably need to see the details to offer suggestions. :slight_smile: