LNJ

LNJ extends the Neighbor-Joining algorithm to construct phylogenies that allow internal nodes to represent extant (live) taxa, enabling trees that include both ancestral and living entities while minimizing total branch length.

Key Features:

• Live Phylogeny Construction: Accommodates live taxa as internal nodes in phylogenetic trees so internal nodes can represent extant organisms.
• Heuristic Approach: Builds on Neighbor-Joining by evaluating, at each step, the traditional join versus an alternative that inserts a live internal node and selecting the option that reduces the sum of branch lengths.
• Versatility Across Data Types: Applies to viral genomes, bacterial genomes, and non-biological datasets such as images and texts by representing content similarity as phylogenetic relationships.

Scientific Applications:

• Viral Genomics: Generates phylogenies for viral genomes that include both ancestral and contemporary taxa to inform virus evolution and epidemiology.
• Bacterial Genomics: Produces phylogenetic hypotheses for bacterial genome data that incorporate living organisms as internal nodes.
• Non-Biological Data Analysis: Visualizes relationships in image and text datasets by mapping content similarity onto phylogenetic structures.

Methodology:

Iteratively selects pairs of taxa (or nodes) that minimize total branch length and, at each iteration, compares the traditional Neighbor-Joining join to an alternative introducing a live internal node, choosing the option that yields a shorter overall tree.