If one decides to create a consensus sequence, then you can use IUPAC ambiguity codes ( ) Individual or one sequence where every base is Either you can create consensus sequences or two separate sequences for each Secondly, there seem to be three options for dealing with heterozygous data from diploid organisms. Nagging feeling that it's an invalid use of that type of data. Those analyses in the same way I use sequences. There's nothing stopping me from throwing the concatenated SNPs into any of Simulations for phylogeographic hypothesis testing, e.g. Use concatenated SNPs in the same way as Sanger sequence data for: divergenceĭating? Bayesian skyline plots? Mutation rate estimation and coalescent It's the biogeographic tests that are giving me pause. Using SNPs just for tree-building - if for no other reason than it seems to be Retaining the invariant sites in RAD data, and not just boiling down to the Linkage, chance, etc.], are giving you the same story of phylogenetic Genetic blocks which, for one reason or another [physical linkage, statistical With only 1-2 SNPs per fragment] could be concatenated and thought of as Similar thought, that loci showing the same topology [which would be VERY basic (Actually, one of my committee members had a Groups of loci to make new matrices is interesting. discussing new methods implemented in STACKS. Julian Catchen also just published a paper in Mol. What kind of biogeographic analyses are you interested in? Many analyses simply require a tree or set of trees along with locality info for your terminals. I personally do not see a difference between concatenating RAD loci for phylogenetics versus concatenating loci obtained from Sanger sequencing. You could then define each of these sets as a 'locus' to perform coalescent-based species tree inference. A second approach, which I have not seen much discussion on, would be to create sets of concatenated loci with enough variation to obtain some phylogenetic resolution. One approach is to concatenate all RAD loci for each individual to create a supermatrix and then run standard concatenated phylogenetic analyses. You must also decide how you wish to incorporate information from heterozygous alleles (using a consensus sequence for example). Because RAD read lengths are still relatively short (~200-250 bp), I think that some sort of concatenation will be required. haplotypes), I think that there are a variety of ways to proceed. If one is interested in building trees using standard methods (i.e. Your questions are andeed valid and I am happy to give my two cents.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |