GTM

GTM merges disjoint subset trees into a cohesive species tree by performing polynomial-time unblended merges that minimize topological distance to a precomputed guide tree for large-scale phylogeny estimation.

Key Features:

• Disjoint Tree Merger (DTM): Implements a DTM approach that merges trees computed on disjoint subsets of species into a single species tree.
• Polynomial-time algorithm: Executes the merging procedure in polynomial time to improve computational scalability on large datasets.
• Unblended merger: Produces an unblended merger by adding edges between subset trees rather than blending or averaging multiple trees.
• Guide-tree minimization: Adds edges between subset trees to minimize the topological distance relative to a precomputed guide tree.
• Pairwise distance matrix: Uses a matrix of pairwise distances between species as part of the merging process.
• Preservation of subset topology: Aims to preserve the topological structure suggested by the guide tree and the initial subset analyses.
• Empirical performance: Reported empirical evaluations indicate accuracy comparable to or better than previous DTM methods with improved computational speed.

Scientific Applications:

• Large-scale phylogenomics: Enables construction of species trees from large phylogenomic datasets assembled via subset analyses.
• Species tree estimation: Facilitates estimation of comprehensive species trees from independently computed subset trees.
• Evolutionary biology: Supports inference of evolutionary relationships among taxa across extensive datasets.
• Biodiversity assessment: Assists in analyses that require scalable reconstruction of species relationships for biodiversity studies.
• Comparative genomics: Provides species-tree backbones useful for comparative genomic analyses across many taxa.

Methodology:

Partition species into disjoint subsets, compute phylogenetic trees for each subset, compute a matrix of pairwise distances between species, use a precomputed guide tree, and add edges between subset trees to minimize topological distance to the guide tree to produce an unblended merger; the algorithm runs in polynomial time.