Typical GC content is around 70 % (Streptomyces). To my knowledge, there is no reliable experimental data on frequencies of different kinds of mutations in this group (except Mycobacterium tuberculosis, but I am not sure as to whether we can extrapolate Mtb data on free-living strains, or can we?). What would the null hypothesis be in this case? Can we gain necessary information through comparison of related genes/genomes?
I would recommend using ModelTest or related programs such as FindModel.
Posada D, Crandall KA. 2001.Selecting the best-fit model of nucleotide substitution. Syst Biol. Aug;50(4):580-601.
Nucleotide composition bias is usually attributed more to bias in selection (growing environment temperature for example with warmer environments favoring GC-rich and cold environments favoring AT-rich genomes) than to bias in mutation frequencies. But even if the total evolutionary process of mutation and selection is not completely modeled by a phylogenetic analysis, the results can still be accurate over fairly large evolutionary distances. At extremely large distances, even the best models will tend to suffer bias grouping GC-rich organisms with other GC-rich organisms due to the composition bias alone.
http://www.hiv.lanl.gov/content/sequence/findmodel/findmodel.html
I reject the null hypothesis in this case. My best guess is that one could potentially extrapolate Mtb data on free living strains and could potentially gain interesting information through comparison of related gene/genomes. Why don't you try a BLAST analysis of the genes that you are interested in comparing and see what you come up with. Look for conserved sequences.
- Badges
- Science topic
- Similar topics
- Biological Science
- Genetics
- Genetic Analysis