There is no a priori reason why a mitochondrial genome should be a good proxy for the evolutionary history of the autosomes. As Balloux pointed out in 2010 (doi:10.1038/hdy.2009.122), mitochondrial genomes are good candidates for selection, and as such should be treated with caution when inferring organism-level phylogenies. Are questions relating to this issue a part of your project?
Thanks for your question, Shaun.
As I agree with you that mitochondrial genomes are good candidates for selection, this fact does not necessarily means that they won't provide good approximations for the evolutionary history of species.
I doubt that anyone would be interested in the evolutionary history of autosomes, as you state. As natural researchers in evolutionary biology, we are interested in the evolutionary history of populations, species and higher taxonomic clades. Data from autosomes can also provide mistaken phylogenies, mainly if they are performed using single genes or a small amount of them. The history of genes and the history of species are frequently different.
This point leads us to something that you must understand: the difference between phylogenetics and phylogenomics! Some people (me included) believe that adding many genes in phylogenies (phylogenomics) will provide better evolutionary histories than studying single genes (phylogenetics). In our mitochondrial phylogenomics project we try to concatenate genes in a supermatrix to provide a more trustworthy phylogeny.
On the other hand, mitochondrial DNA information such as the one presented in Cytochrome B, 16S rRNA and Cytochrome Oxidase (among others) are classical markers to study the evolutionary relationship of clades along the tree of life. They have been used thousands of times to provide some of the most accurate deep phylogenies known in Biology today.
Yes, these questions are not only part of my project but also part of other research programs we develop here in Brazil in our lab (LAMPADA).
Regarding those questions I would invite you to read one of the most comprehensive works on phylogenomics published up to today, a work on which we have had a small participation in the sequencing and analysis of bird genomes. The international avian phylogenomics project sequenced the whole genomes of 48 aves and provided an evolutionary tree produced after the concatenation and analysis of more than 8000 genes, conserved introns and ultraconserved elements (10.1126/science.1253451). Please take a look in the Figure 2 there. In my opinion, this is one of most interesting picture in phylogenomics ever produced and I always show it in my lectures.
Besides, we know that phylogenetic theories and models produced by mathematicians and statisticians behave better in neutral regions of the genome. Thinking about that, our group developed a program that tries to find genomic regions that are as distant as possible to genes, avoiding hitchhiking effects of molecular evolution. We therefore produced a software called ALFie that finds those "anonymous loci" and perform searches in close-related genomes trying to analyze evolutionary aspects in those as-neutral-as-possible loci. The problem with this approach is that we need to have at least one very well known and annotated genome to start the analysis. Please take a look in our Genome Research publication at 10.1101/gr.203950.115.
In any case, any sort of analysis can be biased since the evolutionary history of genes and even neutral regions does not necessarily follows the evolutionary history of populations, species and higher taxonomic clades. Therefore, evolutionary analysis is always inferential and it is hard to know what really happened in the past. We do the best we can using the data we can retrieve in different cases.
Our publications frequently state that we are telling the history of species "as told by" mitochondrial genes and/or genomes. Although these analyses can produce biased results, this can be said for each and every evolutionary analyses. It is probably better to analyze this concatamer of 13 mitochondrial genes in a supermatrix then using a single or a couple nuclear markers that can be biased too. We believe that mitochondria can store relevant genetic information because they are non-recombinant, maternally inherited (in general) and can be expected accumulate mutations in a time-calibrated way.
This way, we do mitochondrial evolutionary analyses when we have mitochondrial data available. We perform concatenation of genes from autosomes when we have these data and study the anonymous loci when we have very well annotated nuclear genomes. All those analyses taken together, using also different methods and statistics, will provide the most reliable inference to understand, discuss and build the most comprehensive history for the evolution of biodiversity on Earth.
A last observation is that, when studying the evolution of biodiversity, DNA data should always be compared to fossil evidences and abiotic evolution, such as the modification of habitats, ecological constraints of river flows, mountain formation and the evolution of continents that provided the division of natural population populations and genetic exchange happening in the past.
I thank you to allow me to explain those points. :-)
Article Whole-genome analyses resove early branches in the tree of l...
Article In silico phylogenomics using complete genomes: A case study...
Thanks for your question, Shaun.
As I agree with you that mitochondrial genomes are good candidates for selection, this fact does not necessarily means that they won't provide good approximations for the evolutionary history of species.
I doubt that anyone would be interested in the evolutionary history of autosomes, as you state. As natural researchers in evolutionary biology, we are interested in the evolutionary history of populations, species and higher taxonomic clades. Data from autosomes can also provide mistaken phylogenies, mainly if they are performed using single genes or a small amount of them. The history of genes and the history of species are frequently different.
This point leads us to something that you must understand: the difference between phylogenetics and phylogenomics! Some people (me included) believe that adding many genes in phylogenies (phylogenomics) will provide better evolutionary histories than studying single genes (phylogenetics). In our mitochondrial phylogenomics project we try to concatenate genes in a supermatrix to provide a more trustworthy phylogeny.
On the other hand, mitochondrial DNA information such as the one presented in Cytochrome B, 16S rRNA and Cytochrome Oxidase (among others) are classical markers to study the evolutionary relationship of clades along the tree of life. They have been used thousands of times to provide some of the most accurate deep phylogenies known in Biology today.
Yes, these questions are not only part of my project but also part of other research programs we develop here in Brazil in our lab (LAMPADA).
Regarding those questions I would invite you to read one of the most comprehensive works on phylogenomics published up to today, a work on which we have had a small participation in the sequencing and analysis of bird genomes. The international avian phylogenomics project sequenced the whole genomes of 48 aves and provided an evolutionary tree produced after the concatenation and analysis of more than 8000 genes, conserved introns and ultraconserved elements (10.1126/science.1253451). Please take a look in the Figure 2 there. In my opinion, this is one of most interesting picture in phylogenomics ever produced and I always show it in my lectures.
Besides, we know that phylogenetic theories and models produced by mathematicians and statisticians behave better in neutral regions of the genome. Thinking about that, our group developed a program that tries to find genomic regions that are as distant as possible to genes, avoiding hitchhiking effects of molecular evolution. We therefore produced a software called ALFie that finds those "anonymous loci" and perform searches in close-related genomes trying to analyze evolutionary aspects in those as-neutral-as-possible loci. The problem with this approach is that we need to have at least one very well known and annotated genome to start the analysis. Please take a look in our Genome Research publication at 10.1101/gr.203950.115.
In any case, any sort of analysis can be biased since the evolutionary history of genes and even neutral regions does not necessarily follows the evolutionary history of populations, species and higher taxonomic clades. Therefore, evolutionary analysis is always inferential and it is hard to know what really happened in the past. We do the best we can using the data we can retrieve in different cases.
Our publications frequently state that we are telling the history of species "as told by" mitochondrial genes and/or genomes. Although these analyses can produce biased results, this can be said for each and every evolutionary analyses. It is probably better to analyze this concatamer of 13 mitochondrial genes in a supermatrix then using a single or a couple nuclear markers that can be biased too. We believe that mitochondria can store relevant genetic information because they are non-recombinant, maternally inherited (in general) and can be expected accumulate mutations in a time-calibrated way.
This way, we do mitochondrial evolutionary analyses when we have mitochondrial data available. We perform concatenation of genes from autosomes when we have these data and study the anonymous loci when we have very well annotated nuclear genomes. All those analyses taken together, using also different methods and statistics, will provide the most reliable inference to understand, discuss and build the most comprehensive history for the evolution of biodiversity on Earth.
A last observation is that, when studying the evolution of biodiversity, DNA data should always be compared to fossil evidences and abiotic evolution, such as the modification of habitats, ecological constraints of river flows, mountain formation and the evolution of continents that provided the division of natural population populations and genetic exchange happening in the past.
I thank you to allow me to explain those points. :-)
Article Whole-genome analyses resove early branches in the tree of l...
Article In silico phylogenomics using complete genomes: A case study...
Shaun, while using mitochondrial genomes for evolutionary inferences we indeed came across cases of positions with significant deviations if Ts/Tv in some positions (see attachment), but their number is too small to affect anything. On the other hand, in the same study we came across several cases of putative tRNA remolding due to mutations in their anti-codon areas. Response to events like that would involve a massive selection although the existing methods searching for selection at molecular level would fail to detect it due to their assumption of code constancy. Thus we may speculate that huge waves of selection could go through mitochondrial genomes but we currently have no means to confirm or reject this hypothesis. BTW there is a long story of detecting the remolding in mitochondria (I attach some papers here), so this event although not quite common, still is to be worried about.
Another bad thing about mitochondrial DNA is that it normally is segregated as a single entity while mitochondrial Ne is 4 times smaller in diploids. Thus it is prone to selective sweeps, transgressions and fake bottlenecks. Still it is nice to use it!
Article Changing identities: tRNA duplication and remolding within a...
Article Towards a comprehensive picture of alloacceptor tRNA remoldi...
Article Evolution of mitochondrial genomes in Baikalian amphipods
- Badges
- Science topic
- Similar topics
- Biological Science
- Evolutionary Biology