TREEFINDER computes phylogenetic trees from molecular sequences. The program infers even large trees by maximum likelihood under a variety of models of sequence evolution. It accepts both nucleotide and amino acid data and takes rate heterogeneity into account. Separate models can be assumed for user-defined data partitions, separate rates, separate edge lengths, separate character compositions. All parameters can be estimated from the data. Tree search can be guided by user-supplied topological constraints and start trees.

There is a model proposer to propose appropriate models of sequence evolution. Confidence of inferred trees can be assessed by parametric and non-parametric bootstrapping, various paired-sites tests (ELW, BP, KH, SH, WSH, AU), also in combination with information criteria, edge support by LR-ELW. There are tools to manipulate trees and sequence data, visualize base frequencies, compute consensus and distance trees, count topologies. Confidence limits and other statistics can be obtained for all numerical results. The program can estimate divergence times by rate smoothing.

A graphical user interface makes the use of TREEFINDER very intuitive. Data files and reconstruction parameters can be chosen interactively, results will be displayed on the screen.

TREEFINDER is programmable in the functional and system independent programming language TL, which offers a wide range of functionality and enables the researcher to create problem specific applications. Parts of the software are written in TL and can be modified by the user.

The Tree Reconstruction Dialog