I want to run a cluster analysis employing categorical morphological traits including all the described species of a genus of plants. My matrix includes both multi-state and unknown values for some traits and I am having some difficulties when trying to obtain the euclidean distances of ordered vs unordered character states.
My goals are:
1- To compare the clusters that will be obtained, with the clades of a molecular phylogenetic reconstruction that included 70% of the described species.
2-To be able to infer (or at least make informed guesses) about the general phylogenetic position, at the subgenus level, of the species that remained unsampled in the phylogenetic reconstruction.
Any recommendations will be very appreciated.
Thank you
The package claddiss has a lot of tools to obtain distances (Euclidean, Gower and other) from morphological matrices coded in nexus. It works with missing data, multistate, ordered and unordered. If you wish I could whip up a script for you.
https://mran.microsoft.com/snapshot/2015-05-04/web/packages/Claddis/index.html
Pl consult an interesting literature
Article Morphological differentiation correlates with ecological but...
Thank you for your replies Dr. Roy. However in the attached citation they deal with continuous variables, so although interesting, this could be of limited use for the kind of analyses I want to carry out.
Under any circumstance I appreciate your recommendation.
Best regards
This could have been useful for you... Darling, E. S., Alvarez‐Filip, L., Oliver, T. A., McClanahan, T. R., & Côté, I. M. (2012). Evaluating life‐history strategies of reef corals from species traits. Ecology Letters, 15(12), 1378-1386.
Maybe you are done with your analyses but the reference above is a good example of how to use categorical data for clustering. Best, GA
The package claddiss has a lot of tools to obtain distances (Euclidean, Gower and other) from morphological matrices coded in nexus. It works with missing data, multistate, ordered and unordered. If you wish I could whip up a script for you.
https://mran.microsoft.com/snapshot/2015-05-04/web/packages/Claddis/index.html
However, you could also use your matrix to obtain a phylogenetic tree using parsimony or Bayesian inference, or try to build a split network using SplitsTree. These might me bore useful for doing a comparison with molecular phylogenies.
Thank you very much for the infromation Dr. Coiro. I would really appreciate if you could share a claddiss script as well, so I could give it a try myself.
Best regards.
Rafael Acuña.
- Badges
- Science topic
- Similar topics
- Biological Science
- Evolutionary Biology
- Phylogenetics