I am comparing phylogenetic data sets in elasmobranchs from GenBank, and am under the assumption that in most species COI will be faster, or have a higher mutation rate, even though both loci are constrained by functional requirements of metabolism, since both are housekeeping genes. But certain authors have suggested that when a COI data set doesn't show differentiation, we should try adding NADH2, but I am not convinced this is a productive direction to take the project.
Hi Brenden
I did a partitioned analysis of mt fragments (not published yet) in order to find out which fragment can be used in phylogenetic data or divergence time estimation. Found out COX1 is much more accurate than ND2. ND2 underestimated but COX1 was closer to the reality. In future will send you the accepted manuscript.
Regards,
I guess it depends what you are trying to accomplish. COI is the more common barcode marker for animals so if that's what you're looking for then go with COI. If you're looking to build a convincing phylogeny for a group then neither one nor two mitochondrial markers is going to cut it, really. What are your options? I mean, is there more sequencing that you can do?
I can speak from my analyses of mtDNA for mammals and for insects and Cone Snails. In each, NADH2 is faster than CO1. COI is a really great gene because much of it is quite conserved so you the data is informative for somewhat older divergence times than ND2. My favorite combination (short of using the whole mtDNA sequence) for most systems is the three gene sequence of ND1,ND2, and COI as you cover a nice range of divergence times.
No single gene is adequate, really, because each mtDNA gene has too little information to get robust results - each gives a randomly different gene.
That is the huge problem with using just COI for "barcoding". Far too small a bit of information for generalizing to the whole nuclear and mt genomes to be useful for separating species. Never mind that it now entrenched as a (bad) method for doing so.
Likewise with a full phylogeny you must use a combination of genes. Ideally you combine all 13 mtDNA genes for a first go. But you must partition the data at least gene by gene and better yet by codon position since a single model of sequence evolution will not suffice to describe the variability among the genes. You will have a biased model.
This paper may have some useful information for you:
Spatial Covariation of Mutation and Nonsynonymous Substitution Rates in Vertebrate Mitochondrial Genomes
Richard E. Broughton, Paulette C. Reneau
Molecular Biology and Evolution, Volume 23, Issue 8, August 2006, Pages 1516-1524 https://doi.org/10.1093/molbev/msl013
The genbank database has tons of elastimobranch complete mtDNA. So should be straightforward to see what the maternally inherited mtDNA has to say about the evolution of this large and important group.
- Badges
- Science topic
- Similar topics
- Biological Science
- Evolutionary Biology
- Phylogenetics