In the litterature I found that we could use at all three different models ( 1. Neighbor-joining trees, 2. Maximum Likelihood and
3. Maximum Parsimony and Bayesian Inference). I would know what is the method that I could apply for my plasmids; knowing that they assigned to the same IncL/M group.
Thank you
All methods are good for different purposes. Maximum Parsimony is nice for identifying synapomorphies in the sequence and is OK for trees if the divergence time is not too deep or if the amount of evolution is moderate or low.
Neighbor-joining is fine for very fast resolution but can be biased because it is algorithmic so does not actually explore tree space so cannot supply support for alternative trees that also may be very good. Hence the tree may differ from those inferred from Bayesian and Maximum Likelihood Trees.
Maximum Likelihood (ML) trees are the most reliable If you have first chosen the best model of sequence evolution by comparing models using Likelihood Scores and/or Akaike Information Criteria (AIC) or Bayesian Information Critera (BIC). ML takes longer than other methods but with a good model will give you the very most robust, lowest variance, least biased results.
Bayesian methods are a maximum likelihood analysis with a slightly different tree search method and with the addition of a term describing prior probabilities. Unfortunately both the tree-search method and the priors can badly bias the tree and the branch length estimates.
A paper from 2012 still describes the benefits and pitfalls of each method: Yang and Rannala 2012. Nature Genetics 13:303
Hi,
All the phylogenetic tree methods are good and it depends on you what you have to analyse from the evolutionary trees.
Generally speaking, in my opinion, Neighbor Joining (NJ) method is the best suitable method for evolutionary tree construction because it takes into account the fact that evolution has occurred and is occurring throughout. It is way better than the UPGMA method for tree construction.
Dear Sir,
I think that you can use any one of the phylogenetic tree methods, it depends on the aim of your analysis. However, Neighbor Joining (NJ) method is the well recommended.
Best regards .
No neighbor doing is not best except for getting a rough first view of the quality of your data an a preview of what you might find with a better method of maximum likelihood. So many advantages to likelihood. Salmi please cite some literature that makes you believe that you think NJ is best. See my earlier response.
- Badges
- Science topic