How To Draw Phylogenetic Tree From Binary Data Getting From Rapd Analysis

Question

[sajinak9]

How can i draw phylogenetic tree from RAPD binary data ( for example , 1 0 1 1 1 1 0 1 1 0 0 1 )

[David W]

Pick software that handles 'discrete' or 'binary' data, find the file format is uses, follow it.

Answer 1

[Josh Herr]

The data you show will not produce a phylogenetic tree, but if you have a matrix derived from RAPD binary data for more than two taxa you will be able to draw a phylogenetic tree (you may even be able to compute one by hand with such a small matrix!). You can draw a phylogenetic tree from (virtually) any kind of data in matrix form. Here's a link for a good introductory course handout on how to compute a phylogenetic tree. Bacterial Phylogeny to a similar question on how to get started with making a phylogenetic tree.

Answer 2

[aidan-budd]

Some software that accepts alignments of binary characters as input data in includes:

• PHYLIP - this set of programs take that kind of data http://evolution.genetics.washington.edu/phylip/progs.data.disc.html
• MESQUITE
• RAxML Phylogenetic tree with DNA, binary and multi data - RAxML UPDATE

[Josh Herr]

+1, These are all great examples of software for phylogenetic analysis, I would also add PAUP (old but classic) and MrBayes as rounding out essential programs. By nature, all phylogenetic programs accept binary character data, so any program -- as long as it works -- is acceptable.

[aidan-budd]

Yes, absolutely, PAUP and MrBayes are definitely two other packages to turn to here.

As Josh says, other software, such as PhyML, can be also be run with binary data. However, again taking the example of PhyML, it doesn't allow you to explicitly choose a model for this kind of data (at least as far as I'm aware for the versions I've worked with), you rather have the choice only between amino acid or nucleotide sequence models. Thus you have to make a bit more effort and thought to put binary data into it (you could code 1 as "T", 0 as "G", for example, and run using nucleotide models).

What makes the PHYLIP programs listed above, MESQUITE, RAxML, MrBayes, and PAUP* particularly relevant here is that they explicitly implement models for binary data of this kind e.g. you can load binary data into RAxML and specify using the -m option the model BINGAMMA (binary + gamma).