I'm working on an interespecific phylogeny of psbA-trnH haplotypes and I would like to run Bayesian and Maximum Likelihood analyses considering gaps as fifth base. There are some indels blocks specific to each species, then I think this information is important in the separation of the three species involved in the analysis. Could anyone tell me which software can I use to run a Maximum Likelihood analysis considering gaps as fifth base?
I do this by using RAxML. The suggestion from the author was:
"If you have a DNA alignment you need a script to transform the characters A -> 0, C->1, G->2, T->3, - -> 4 and then run a multi-state based analysis with a GTR model which will treat "-" as 5th state. We have tested this a little bit and tree accuracy seems to improve slightly. Nonetheless, this is only a very crude way of treating indels in likelihood computations and there do exist more elaborate approaches.
You can try POY 5 and Baly-Phy as well
www.biomath.ucla.edu/msuchard/bali-phy/
http://www.amnh.org/our-research/computational-sciences/research/projects/systematic-biology/poy/download
Best
max
Yoannis
Simply Paup or Garli, you just have to declare a "new state" for the "fith" base. If you've got common indels within or between species, it is often better to consider it as a single "mutation" (as it is a unique event).
Cheers from Luc
here a example on paup:
#NEXUS
begin data;
dimensions ntax=24 nchar=936;
format datatype=dna gap=- interleave;
options gapmode=newstate;
matrix
Hi Luc,
Actually, I used PAUP software, and I followed your recommendation using gapmode=newstate in my infile, but It was not work!!! I noted it because I have a sequences that have differences only in 1 or 2 indels, and when I obtained the distances between sequences I got 0 values! :(
I think the option to use the gaps as a fifth character in PAUP is meant to be used with parsimony methods. ML and Bayesian methods ignore gaps.
- Badges
- Science method
- Similar topics
- Biological Science
- Evolutionary Biology
- Phylogenetics
- Phylogenetic Analysis