hello everyone ! I got a question about trait evolution analysis.
I have COI (Cytochrome oxidase I) DNA sequences for all species, and I have excel sheet for all morphological characters for all species eg. (thorax volume, proboscis length, wings and legs length, Max. width of tergite and sternite, mature egg length and number of mature eggs in ovary.........ect.). I need to use R software (ape and caper package).
Now, I want to analysis these data to study trait evolution history (Mapping these traits on the tree, the evolution relationships between these species). So, how could I achieve that goal ?
thank you all ! Any comments would be appreciated !
actually, I am a beginner in this field, I didn't have a lot experience. I need to now that:
1- I should build the tree from different software (like MEGA) and then import this tree to R.
2- Can I import the sequences to R and build the tree from it with out use other software?
3- In which kind of format I should save the sequences to import it to R?
4- how can I link this morphological data to the tree?
Thank you for all.
Aram, when you say that you want to "analysis these data to study trait evolution history" by "Mapping these traits on the tree" you are predefining the relationships according to the sequence data as if the morphology carried no valid phylogenetic signal. Its a very widespread practice, but has no scientific validity. You could compare the sequence and morphology trees and see where they match. Where they do not match you have a problem for which there is no objective recipe to solve.
Well, I am also interesting in doing this some day with my animals, but have some experience only in building trees so far. So my suggestions would be to build a tree in MrBayes using algorithm in any random paper (it should include identification of best evolutionary model and partition combination; then you you this partition and model in settings of MrBayes run). It's good idea to build let's say RAxML of PhyML maximum likelihood tree to check if they will produce similar topology with MrBayes. I would not use Mega NJ tree or tree build in R - seems this is fast and dirty ways to assess topology of the tree, not for reviewers. Here my expertise is over, read manuals and vignettes. As a rule they are good enough to be reproduced. Good luck!
Hello, first, I'm not familiar with R, but I think some software about trait reconstruction maybe help. As I know, two software could do this:
1) Mesquite, you could read the manual book for more info, as I tried, it has a function to trace the trait evolution history along the tree.
2) BayesTrait, this is a terminal command software , you can check the manual book for more details you need.
Hope it helps ! have a good day!
Yanshi
Aram, when you say that you want to "analysis these data to study trait evolution history" by "Mapping these traits on the tree" you are predefining the relationships according to the sequence data as if the morphology carried no valid phylogenetic signal. Its a very widespread practice, but has no scientific validity. You could compare the sequence and morphology trees and see where they match. Where they do not match you have a problem for which there is no objective recipe to solve.
Hi, I think that estimate the phylogeny tree from software instead using R packages is a wise choice, you can using MrBayes, BEAST and other software. As for , inputting a tree from BEAST to APE or Diversitree, you can use a newick format ,and there are R packages can be used to you study , you can employ Geiger ,diversitree, phytools and so on. Due to you measured many characters , I advice you can use PPCA(Phylogenetic Principal Component Analysis) firstly. This will reduce you axis
hope this will help.
best
Sun Lu
Of course, to interpret a tree as representing evolutionary relationships you need to root it using appropriate outgroups (also scored for the molecular and phenotypic dataset), and show evidence for monophyly of the ingroup 'genus'.
thank you John,
I know that I need to have the out group, but my problem is that how can I interpret these morphological character, or how can i connect these characters with the tree? how can I produce data matrix from this excel sheet morphological characters?
If you know, explain it, I can able to send you the data by email to see it.
I get the impression from your latest questions that you do not have access to any systematics expertise at the University of Nottingham or that they do not have courses on the technical aspects of tree building? If that is the case perhaps another university would provide such courses. The details of tree-building and character matching are more than could be effectively explained in this venue, although someone else might be able to direct you to on-line material. Your question about matching morphological characters to a molecular tree indicates that you regard morphology as phylogenetically uninformative and that it only has meaning given to it by the congruence or incongruence with the selected molecular tree. A common practice but without empirical merit. Why even bother?
To Mesquite you can import your Matrix characteristics database and construct your trees (is very easy to use it, you just need to select the tree method you want to run and thats it, there is a lot of manuals you could read). In R you can import your sequences like .FASTA sequence using the ape package read.FASTA command, but I strongly recommend to use other softwares (MEGA is the easiest to use, but if you're going to publish, don't use MEGA 7 because its in a BETA stage), read this webpage, you'll get better understanding of softwares and which one works for what kind of analyses, and if will work for what you want:
http://evolution.genetics.washington.edu/phylip/software.html
Take care!
- Badges
- Science topic
- Similar topics
- Biological Science
- Evolutionary Biology
- Phylogenetics