DecentTree

DecentTree implements distance-based phylogenetic tree construction using the Neighbour-Joining algorithm and its variants, providing optimized and parallelized computation for large-scale sequence datasets.

Key Features:

• Algorithms: Implements Neighbour-Joining and its variants for distance-based phylogenetic inference.
• Optimizations: Provides algorithmic optimizations and parallelized implementations to accelerate tree construction.
• Implementation: Available as a cross-platform C++ library and as a stand-alone command-line application.
• Input formats: Accepts Phylip-formatted distance matrices and FASTA alignments.
• Output format: Produces phylogenetic trees in Newick format.
• Scalability: Designed to scale to datasets exceeding 10,000 sequences, with benchmarks on tens of thousands of genomes.
• Integration: Can be incorporated into other phylogenetic software, exemplified by incorporation into IQ-TREE.
• Performance: Demonstrated faster performance than BIONJ, Quicktree, FastME, and RapidNJ, reaching up to sixfold speedups versus RapidNJ in large-dataset benchmarks.
• Benchmark example: Performance evaluated on a benchmark of 64,000 SARS-CoV-2 genomes.

Scientific Applications:

• Phylogenetic reconstruction: Construction of evolutionary trees from distance matrices or sequence alignments.
• Large-scale genomics: Analysis of very large sequence datasets such as tens of thousands of genomes.
• Outbreak genomics: Rapid tree building for pathogen surveillance and outbreak-scale datasets (e.g., SARS-CoV-2).
• Method benchmarking: Comparative performance evaluation against other distance-based tree-building implementations.

Methodology:

Uses distance-based methods (Neighbour-Joining and its variants) with algorithmic optimizations and parallelization; accepts Phylip distance matrices and FASTA alignments and outputs trees in Newick format.