Open SilasK opened 3 years ago
You are correct that it is often not possible to obtain full-length 16S assemblies from metagenomes. There are many reasons for this, however, all of them can be reduced to the single word: "repeats". Ribosomal genes exist in multiple copies in the genome, making assembly difficult. This difficulty is extremely amplified by the presence of close species / strains in the metagenome as the repeats become interspecies. Certainly MAGs will miss rRNA contigs as binning is only done for contigs longer than 1.5 - 2 kbp and the rRNA contigs will certainly be shorter. Proper binning refining procedures could help here.
There is no safe and reliable way to assemble complete rRNA genes due to fundamental limitations of short reads (e.g. fragment length that does not span the repeat in question, etc). I'm afraid many external "pipelines" are prone to produce chimeric results. That said, I'm not sure what could be the "special attention" here :)
If I metal spades can assemble 16SrNA which then are binned into high-quality bins, would you trust these sequences? Obviously you could do som checks if it's only a contig with 16S gene and if the Taxonomy mages or so. But in the case where you have a rNA gene inside a longer scaffold?
I got a reviewer, tat sayd you can't trust them.
Is your feature request related to a problem? Please describe. For generic questions use Q&A section in the Discussions forum above.
rRNA and tRNA are thought to be difficult to be assembled from metagenomes, due to their similarity. I found that metal spades do already a good job in assembling some of them. However, I'm not sure about their quality
Describe the solution you'd like
Ideally, I would like to search for rRNA (and tRNA) during the assembly of metagenomes and then pay special attention to their assembly (using normal spades)
Describe alternatives you've considered
I saw phyloFlash which does more or less this. But I think it would be ideal to have it inside the spades workflow.
Additional context