scale bar meaning

[Kevin Theis]

Hi all,
I have made a RAxML tree illustrating the phylogenetic relationships
between bacterial operational taxonomic units (OTUs) from
environmental samples and neighboring bacterial type strains. In
microbial ecology, operational taxonomic units are defined exclusively
upon percent nucleotide similarity, so the exact meaning of the scale
bars on trees is important with respect to interpretations of
phylogenetic similarity. Specifically, scale bars can be very
misleading if interpreted incorrectly. I have found that many
researchers define the scale bar in a RAxML tree as an estimation of
the number of nucleotide substitutions per site. This doesn't seem
correct to me. It is reflective of estimated evolutionary distance,
but it doesn't seem that this translates directly to percent sequence
similarity. I am currently defining the scale bar as follows: "The
scale bar reflects the likelihood of evolutionary change per
nucleotide site as inferred by the RAxML algorithm." Is this correct?
I welcome any suggestions and clarifying comments. I have consulted
the manual and searched this user group, but if I missed prior
discussion of this matter I apologize.
Thank you for your time,
Kevin Theis
Michigan State University

[Alexis]

Dear Kevin,

The scale bars, or branch lengths to be more precise, correspond to:

"the mean number of nucleotide substitutions per site" on the
respective branch.

Joe Felsenstein has written an excellent textbook "Inferring
Phylogenies" that explains the differences between
likelihood-based distances and simple percent sequence
(dis-)similarity.

Alexis

[Kevin Theis]

Thank you very much Alexis, for both the clarification and the
recommended text.
Kevin