mtDNA is different from X and Y chromosome DNA.

mtDNA is DNA from the mitochondria. It tracks the maternal line. Mitochondria are little organelle's within cells which have a lot to do with providing energy for the cells functions. The theory goes that way back in the mists of time, one bacteria ate another bacteria but didn't digest it. The "eaten" bacteria discovered that living inside another bacteria had advantages. The "eater" bacteria discovered that having this particular other bacterium inside itself also conferred some advantages. Thus, the advent of the mitochondrion.

Whatever the case, mitochondria have their own DNA, which has nucleotide codes for molecular processes that occur inside the mitochondria. Many of those processes have to do with dealing with an important biological molecule called ATP. ATP is the "powerhouse molecule" of the cell. It's the molecule that cells use to power the work that they do. Sperm SWIM, don't they? That's a lot of work, isn't it? Sperm have a LOT of mitochondria. However, the sperms mitochondria are not added to a fertilized oocyte. That oocyte only contains maternal mtDNA.

You see, during the process of fertilization, the sperm attaches to the oocyte (the egg) cell membrane and injects chromosomal DNA. However, it ONLY injects chromosomal DNA, no mitochondria are injected. Therefore, the mtDNA is passed on from generation through generation through the mother, essentially unchanged except for the random and exceedingly rare process of mutation.

Different human population groups contain mitochondria which are somewhat distinct from each other, in terms of the various forms of the genes inside them. These "various forms" are call polymorphisms and are represented by subtle variations in the order of four molecules that Forrester Modern talked about: Adenosine, Thymidine, Cytosine and Guainidine.. Thus, an/some ancestral group(s) of people might be identified, by working backwards from current arrays of mtDNA polymorphisms. There are computer programs and fancy algorithms which use the principle of maximizing parsimony (parsimony = the simplest answer is the most likely answer) which can make an educated guess at what these ancestral polymorphisms would be like, by choosing the simplest possible series of mutational events that might result in the modern array of polymorphisms.

In fact, since there are estimates of the average rate of mtDNA mutation, some researchers even try to pinpoint general times when different polymorphisms occurred, thus putting some sort of timeline on the human mtDNA family tree.

To simplify...If Panache and Alan H share 12 polymorphisms in their mtDNA....and Panache and McMurdo share 8 polymorphisms, then Panache and Alan H are more closely related than Panache and McMurdo.