Hi Reza,

It depends somewhat on what you're trying to get at. Obviously the longer the gene, the more sequence data you'll have for resolving their relationships, but it also depends on how long ago the species diverged from each other. Mitochondrial genes are generally fast evolving, and can be very useful for inferring relationships of recently diverged species/populations. So if you're dealing with recently diverged species, you'll want to pick the quickest evolving of those genes. If you're looking at early divergences, you'll potentially want a more slowly-evolving gene since much of the phylogenetic signal could be washed out if it evolves to quickly.

I would suggest going onto Genbank and seeing which of the four genes is available for various snake species. Whatever is up there, make an alignment for each gene and estimate a gene tree for each. The resulting branch lengths should give you an idea of the evolutionary rates for each gene.

-Chris

Hi Reza, 

It will depend on what hypothesis you are trying to  test. But regardless, I just want to point out that using mtDNA genes you will not have a phylogeny. You will have a genealogy, or a female family history. The results for "phylogeography" cannot be interpreted as an evolutionary history of a species or a genus, because it does not take in consideration the males (male dispersal, nuclear gene flow etc.). And there are several cases where female haplotypes (from mtDNA) do not agree with nuclear gene flow.  So be very careful on what you are using mtDNA for, because most of the good journals will not publish your research if you use mtDNA to infer evolutionary relationships. 

The mitochondrial coding genes are all similar in terms of phylogenetic signal and mutational pattern. ND4 has greater representation in Genbank, although ND2 has also been used extensively.