search

tseemann / barrnap

:microscope: :leo: Bacterial ribosomal RNA predictor
GNU General Public License v3.0
221 stars 40 forks source link

Support plant mitochondrion #2

Open sjackman opened 10 years ago

sjackman commented 10 years ago

Plant mitochondrion have three rRNA genes:

sjackman commented 10 years ago

See Cycas taitungensis mitochondrion.

sjackman commented 9 years ago

Bump?

tseemann commented 6 years ago

Not sure I can do this, as rrn26 is spliced into 2 exons in some plants?

# 2 exons
     rRNA            complement(join(146852..148823,149542..150368))
                     /gene="rnl"
                     /locus_tag="MapooMr01"
                     /gene_synonym="rrn26"
                     /product="large subunit ribosomal RNA"

# 1 exon
     rRNA            complement(126101..129506)
                     /gene="rrn26"
                     /locus_tag="CQR31_mgr02"
                     /product="26S ribosomal RNA"
                     /db_xref="GeneID:35198937"
sjackman commented 6 years ago

Can you sweep that under the rug and identify only the single exon genes for now? That'd be better than what I have now. Either identifying spliced genes as partial genes or not at all would be okay. I'm using other pipelines in parallel to try and identify the spliced genes.

sjackman commented 5 years ago

Hi, Torsten. I've found that barrnap --kingdom bac identifies plant 18S and 26S rRNA just fine. It does not however detect plant mitochondrial 5S rRNA (at least in my conifer species). RFAM 5S_rRNA (RF00001) with Infernal detects it just fine though!

Would you consider adding this RFAM family to Barrnap?

You can download an alignment of the family here so that you can build an HMM for hmmer. http://rfam.xfam.org/family/RF00001#tabview=tab2

You could also download the prebuilt Infernal covariance model (CM) here: http://rfam.xfam.org/family/RF00001#tabview=tab9

The covariance model includes an HMM within the model. Search for the line HMMER3/f [i1.1 | October 2013]. I don't know whether the file format is compatible with hmmer. If it is, you could just extract this HMM from the covariance model.

Alternatively, would you consider running infernal as an option instead of hmmer? It scans first with an HMM and then checks the hit against the RNA covariance model (at least as I understand it).

5S_rRNA (RF00001) is a member of clan 5S_rRNA (CL00113) which includes only one other family, mtPerm-5S (RF02547). I don't need the permuted form, but thought I'd mention it. http://rfam.xfam.org/clan/CL00113 http://rfam.xfam.org/family/RF02547

The permuted mitochondrial 5S rRNAs in brown algae represent the most unconventional case, where the closing helix I, which otherwise brings together the molecule's 5′ and 3′ ends, is replaced by a (closed) hairpin resulting in an open three-way junction.

https://en.wikipedia.org/wiki/5S_ribosomal_RNA

Here are the RFAM bacterial family and clans for rrnS and rrnL. The clans include the families for other kingdoms. The bacterial families identify plant mitochondrion rrnS and rrnL.

SSU_rRNA_bacteria (RF00177) http://rfam.org/family/RF00177 SSU (CL00111) http://rfam.org/clan/CL00111

LSU_rRNA_bacteria (RF02541) http://rfam.org/family/RF02541 LSU (CL00112) http://rfam.org/clan/CL00112