Weighbor

Weighbor reconstructs phylogenetic trees from distance matrices using a weighted neighbor-joining algorithm that downweights longer distances to reduce their influence on tree topology.

Key Features:

• Weighted Distance Integration: Applies weights to distance estimates to reduce the influence of longer distances, accounting for increasing estimation error with distance.
• Likelihood-Based Criterion: Models distances as correlated Gaussian random variables with means and variances derived from a probabilistic model of sequence evolution and evaluates joins using a likelihood-based criterion.
• Dual-Term Criterion: Makes joining decisions using an additivity term that evaluates deviations from additivity in implied external branches and a positivity term that assesses confidence in the positivity of internal branch lengths.
• Efficiency and Accuracy: Produces phylogenetic trees that are qualitatively and quantitatively similar to maximum-likelihood reconstructions while offering faster computation.
• Robustness Against Common Pitfalls: Reduces errors associated with long-branch attraction and long-branch distraction relative to neighbor joining, BIONJ, and parsimony methods.

Scientific Applications:

• Evolutionary biology: Reconstructing relationships among species or sequences to infer evolutionary histories.
• Genomic studies: Building distance-based phylogenies from large genomic datasets where computational efficiency is required.
• Comparative genomics: Inferring phylogenetic relationships to support comparative analyses across genomes.
• General phylogenetic analysis: Producing robust distance-matrix-based trees for diverse research fields requiring phylogenetic inference.

Methodology:

Implements a weighted neighbor-joining algorithm that applies weights to distance estimates and uses a likelihood-based criterion modeling distances as correlated Gaussian random variables (with means and variances from a probabilistic model of sequence evolution) and evaluates joins using additivity and positivity terms.