I want to analyze mutations in the core-genome of Xanthomonas oryzae, a bacteria species that comprises two pathovars: pv. oryzae and pv. oryzicola. I already have the core-genome identified (those genes conserved in at least 95% of the strains in each pathovar), and now i want to identify those genes that are less conserved. To do it i am thinking in calculate some conservation score for each gene based on the multiple sequence alignment by measuring the probability of each aminoacid occur in each position in both pathovars and then extract the mean shannon entropy. The Shannon entropy would be used to as the main measure of the conservation of the gene, and would be calculated for each pathovar-gene and then a mean . Thus, with a score for each gene, i would calculate the z-score and select those with outlier z-score and perform a GO enrichment analysis on them. The idea is to identify genes that are differentially variable or differentially conserved in the pathovars and the species.
Can anyone give me some advice or say if this method is appropriate?
It sounds like you are saying you have sequenced the complete genomes of a few hundred isolates of each of the two pathovars. That is a lot of data. It is possible to measure diversity of each gene within pv. oryzae and within pv. oryzicola and compare the two levels of diversity. Entropy measures the diversity without taking phylogenetics or evolution of each pathovar into account, or you could measure rate of evolution of each gene which would take the phylogeny into account.
Is there no horizontal transfer of genes between these two pathovars? If not, why are the names as pathovars of the same species rather than being named as two different species? Genes move in between Escherichia, Salmonella, Shigella and other enterobacteria all the time, so it seems like you should check for horizontal transfer of genes as one mechanism of different apparent rates of evolution.
It may also be worth looking at smaller or larger regions of the genomes than genes. Some genes have hypervariable and conserved regions so that the average rate of evolution over the gene does not reflect how the gene is evolving.
It is also possible that for your purposes, you do not care what mechanism is causing diversity (horizontal transfer, hypermutation, strong selection etc) and you only care about how much diversity there is. In that case there is not much point in taking the trouble to look at evolution.
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Bioinformatics