1. High mutation rate due to limited repair system (5-10 times that of nuclear DNA).

This is because mitochondria are essentially bacteria (prokaryotes) living within eukaryotic cells. Bacteria have less DNA damage and repair machinery than the average eukaryote, and mitochondria have even less because they have lost most of their genes. The mitochondria import a lot of proteins and metabolites that they need from the eukaryotic cytoplasm. but they don't get all the DNA repair enzymes.

2. Conserved structure of the genome.

It is difficult to create a multiple sequence alignment when the gene order is different between the organisms being studied. Over time, bacterial genomes tend to get scrambled with chromosomal inverstion, insertions, deletions, horizontal transfer by phages, etc. The mitochondiral genomes tend to be rather stable within a given group of related eukaryotes such as the mammals, so it is easy to align them for phylogenetic analyses.

3. Lack of recombination.

Bacterial DNA (and eukaryotic nuclear DNA) is full of recombination events and this confounds phylogenetic reconstruction. Humans don't recombine with horses or frogs of course, but we did rather recently receive alleles from Neanderthals after being separated from that lineage for a long time, and when we split from the chimpanzees and gorillas there was incomplete lineage sorting which means that different genes in the human genome have different histories.

4. High phylogenetic signals of COI (cox1) than other genes of mitochondria.

Different genes, and even different regions of a single gene, evolve at different rates. Site that evolve quickly are good for seeing the difference between individuals of a species or between subspecies, but not so good at deep relationships (comparing mammals to fish for example) because too much divergence leads to more "noise" than "signal" in the data. Highly conserved regions are better for studying the deeper relationships.

COI genes are species specific

Thanks all Dear colleagues.

COI sequences are neither species specific nor essentially different from the mtDNA genes.

They are useful because they are long and generally have a nice combination of conserved stretches and faster moving stretches.that allow analysis with broad range of evolutionary rates.

This is one of the the most widely used genes for that reason and also simply because early studies

No gene is species specific Evolutionary processes proceed without "knowledge' of human imposed species boundaries.

!. .The higher mutation rates of mtDNA compared to nuclear genes may be due to relaxed repair mechanisms but also to the nature of the 13 protein-coding genes that are involved with metabolism.

2 Gene order is often pretty well conserved across families with orders and even classes of many organisms (notably Eutherian mammals) but are definitely not conserved across larger taxonomic levels.

3. Because the mitochondria are maternally inherited they do not interact with paternal genes generally speaking - his no recombination of genes between parents are happens with nuclear genes which come from both the mother and the father and sort out differentially in the offspring.

4. COX1 is not among the faster evolving mtDNA genes. (see first few paragraphs).

Maybe this is too obvious too mention but, the reason #1 is often mentioned is that mitochondrial sequence variation is greater than nuclear variation, making analyzing mitochondrial variation valuable for closely related organisms. This is often mentioned under, "Why use mtDNA?" and includes other reasons like 1) the fact that since there are numerous copies of each mt segment in every cell (several per mitochondrion), it is easier to amplify a mt than nuclear segment if the tissue is degraded (ancient DNA studies), and 2) mtDNA is maternally inherited and can give highly conserved info concerning the maternal line of a sample (as mentioned above)(lack of recombination, like the Y chromosome).

I agree completely with David. But I have had oddly wonderful results for Cone Snails and for Eutherian Mammals back to the early Cenozoic (late Paleoicene to Eocene and later) .Of course this is not early for many systems but for many crown species today it is excellent. So much cleaner data than nuclear for this time period. In general much less noise, more signal.

Thanks all Dear colleagues

Really interesting answers.

Dear Dr. Drovetski, so you mean using COI as a standard barcoding region is wrong ?!

I think, it might be true, because in biology we are dealing with alive organisms which are the results of evolution, not computers!

What about whole mtDNA sequencing (Mitogenomics)? Will this help us to differentiate species easily and make reliable phylogenetic trees? Or still, there will be gaps and problems?

Thanks Dear Dr. Drovetski

I really appreciate this. Very useful information.