NEXUS Format:

The CIPRES Science Gateway supports Nexus blocks are supported natively by a given program. In most cases, command blocks in Nexus files are the preferred way to conduct analyses. Nexus files are difficult to parse because there are many format variations in use, and many different ways to encode the information. The tool Format Validator is provided to help with this. Mark Holder created this tool for the CIPRES group, based on the NCL program of Paul Lewis (also found in Phycas). While philosophically we favor more strict standards, each program here has its own philosophy and capabilities for Nexus interpretation. Accordingly, the documentation for each indvidual application should be the resource of first resort for issues that arise.

The original document specifying Nexus file formats is here [pdf].

IMPORTANT NOTE: Normalized Nexus will not be recognized by MrBayes 3.1.2

Normalized Nexus headers divide taxa and characters into sepoarate groups, look like this:
begin taxa;
dimensions ntax=12;
taxlabels a b c d e f g u v x y t; end
begin characters;
dimensions nchar=432;
format datatype=dna missing=?;

MrBayes 3.1.2 expects characters and taxa to be in a single data block, like this:

begin data;
dimensions ntax=12 nchar=432;
format datatype=dna missing=? gap=-;

This can be corrected manually with a text editor.

If your NEXUS file is rejected please also try the Format Validator tool for help. Another resource is a list of formatting issues that we have run into from submissions to CIPRES Portal v 1.0. They can be found here, along with solutions.

If there is a tool or a feature you need, please let us know.


CIPRES – Cyberinfrastructure for Phylogenic Research