FastME is a software package for the fast and accurate inference of phylogenetic trees from distance matrices. It implements algorithms based on the Balanced Minimum Evolution (BME) principle, a distance-based criterion closely related to the Neighbor Joining (NJ) method. The goal of the BME framework is to identify the phylogenetic tree that minimizes the total estimated evolutionary distance among taxa.

Compared to classical Neighbor Joining approaches, FastME improves tree accuracy through efficient topological optimization procedures. After constructing an initial tree, the program refines the topology using rearrangement operations such as Nearest Neighbor Interchange (NNI) and Subtree Pruning and Regrafting (SPR). These optimization steps allow FastME to explore alternative tree topologies and identify trees with shorter total branch lengths while maintaining a computational cost comparable to NJ.

FastME supports multiple types of input data, including DNA sequences, protein sequences, or precomputed distance matrices, and provides several methods for evolutionary distance estimation. The software also includes features such as bootstrap analysis for branch support, parallel computation, and several tree refinement strategies.

Thanks to its combination of speed, scalability, and accuracy, FastME is widely used for phylogenetic reconstruction, particularly for large datasets where likelihood-based approaches may be computationally demanding.

Reference: Lefort V., Desper R., Gascuel O. (2015). FastME 2.0: A Comprehensive, Accurate, and Fast Distance-Based Phylogeny Inference Program. Molecular Biology and Evolution, 32(10), 2798–2800.