STELAR

STELAR estimates species trees from collections of gene trees by maximizing triplet agreement under the multi-species coalescent model to address gene tree incongruence caused by incomplete lineage sorting (ILS).

Key Features:

• Statistical consistency: Provides statistically consistent species tree estimates under the multi-species coalescent model.
• Triplet-consistency optimization: Estimates species trees by maximizing agreement among triplets extracted from input gene trees.
• Constrained Triplet Consensus (CTC): Formalizes species tree inference as the constrained triplet consensus (CTC) problem.
• Dynamic programming solver: Solves the CTC problem using an efficient dynamic programming algorithm.
• ILS handling: Explicitly models gene tree incongruence due to incomplete lineage sorting (ILS) through the coalescent framework.
• Scalability and accuracy: Designed for scalability and reports high accuracy on large datasets.
• Empirical validation: Demonstrated comparable performance to ASTRAL and improvements over MP-EST and SuperTriplets on simulated and real biological datasets.

Scientific Applications:

• Species tree estimation with ILS: Inferring species relationships from gene tree collections when incomplete lineage sorting affects gene tree topologies.
• Phylogenomic analysis of large datasets: Analyzing large-scale genomic or phylogenomic datasets to reconstruct species trees.
• Method benchmarking: Comparative evaluation and benchmarking of coalescent-based species tree inference methods using simulated and empirical datasets.

Methodology:

Maximizes triplet agreement among input gene trees by formalizing the constrained triplet consensus (CTC) problem and solving it with an efficient dynamic programming algorithm to produce species tree estimates consistent with the multi-species coalescent model.