Horizontal gene transfer (HGT) and some other phenomena, such gene duplication/divergence create significant differences in tree topologies among functional gene and 16S rRNA gene phylogenies (Jones et al., 2008). I have seen in the papers doing Next Generation Sequencing methods, between 2-5% cut off was used to assign DNA sequences to OTUs when they were studying functional genes. I am wondering what this cutoff means when due to HGT (for example), functional genes transfer among different species from one species to another. The same question comes up in the other techniques like DGGE. In this technique the principle is different sequences denature at different denaturant concentrations which results in a pattern of bands and each band represents one OTU. The question is if we can really consider each band as an OTU in the case of functional gene investigation?
Thank you in advance,
The main point, as I understand was "what percentages should be used for functional OTUs distinction"? My answer will be: each gene/protein has its own number if it could be determined at all.
16S OTUs are based on high scale comparison of 16S identities and genomic DNA hybridizations (in vitro or in silico) of the same organisms. It appeared, that 96% of ANI (or 70% of in vitro genomic DNA hybridization) ~ 97% 16S identity => since 70% of hybridization was initially chosen for a species border, the 97% of 16S rRNA gene identity became a species border too. Thus, all this is based on huge amount of data, coming from comparison of genomic DNA and this particular gene. Indeed, you can use few functional genes/their products for phylogenetic analysis in some particular cases (eg. McrA, RecA, RpoB), but to link these phylogenies with an assignment of real taxonomic units you will need the same comprehensive comparison of identities of these genes/proteins and the genomic DNAs of the same organisms (or with 16S since we know the border for 16S). I`m not sure, but probably, for McrA or may be for RpoB thess kind of works were done before.
As for HGT, duplications etc. this is a real problem but also it is correct for 16S (16S rRNA gene, actually, are often duplicated and even multiplicated). All, what I said before is right only for functional genes, suitable for phylogenetic analysis, what means they are also characterized by low HGT and duplications/deletions rates.
- Badges
- Science method