Phylogenetic MetricsThere are a lot of ways to create phylogenetic trees (cladistics, fossil layer ordering e.g.). There are also several ways to create them, by using genetic material alone. Using Gene-family comparisons seems most trustworthy. The probability that the results of a study are coincidence, because only isolated mutations or only two species are compared, does decrease. Such Gene-family comparisons can have different metrics. Molecular phylogenetic techniques investigate:
- Similarity between nucleotide sequences.
- Distance between genes on the genome.
- Similarity in degradation of genes to pseudogenes, and the way they degrade.
- Similarity in synonymous and nonsynonymous* ratio divergences. (With this not only pseudogenes can be recognized, but they can be used IMHO as metric itself too.)
You are probably able to list many metrics more. If anyone possesses a comprehensive overview, I'd like to see it... This is however not exactly the content of this thread.
MathematicsMathematical tools are used to find for example the fewest steps to "come from" a certain nucleotide sequence "to" another. This is also interesting stuff (using Markov Models etc), but neither the topic of this discussion.
Question BackgroundThe metrics I listed differ in their convincing abilities. The phenotype across species corresponds to a certain distribution of characteristics. It is likely to have a genetic code distribution that is likewise (also for someone that beliefs in "design", whatever that entails). However, metrics like the distance between genes are neutral in that regard. There is no functional reason to design the system like that. Moreover, metrics like the gene corruption to pseudogenes that show a similar distribution, are pretty convincing in this context. There is no reason whatever to design the system with that kind of corruption.
QuestionMy question is therefore: Do you know studies with gene families, or at least across several species, that investigate the correspondances in decay of genetic material?
Example: See especially table 1 in
Human specific loss of olfactory receptor genes, Gilad et al.[PDF]
*involving mutations that change the amino acid for which the new codon codes