SnappNet

SnappNet infers phylogenetic networks from biallelic markers using a Bayesian extension of the multispecies coalescent to model reticulate evolutionary events such as horizontal gene transfer, hybridization, and introgression.

Key Features:

• BEAST 2 integration: Operates within the BEAST 2 platform for Bayesian evolutionary analyses.
• Foundational packages: Builds upon Snapp (Bryant et al., 2012) and SpeciesNetwork (Zhang et al., 2017).
• MCMC operators: Leverages novel MCMC operators from SpeciesNetwork to navigate network space and includes Snapp-specific operators for population sizes and mutation rates.
• Model extension: Extends the multispecies coalescent (MSNC) to phylogenetic networks that include reticulate nodes.
• Likelihood computation: Computes the likelihood of biallelic markers sampled along genomes that have undergone reticulation.
• Bayesian framework: Performs Bayesian network inference via MCMC under a shared evolutionary model.
• Comparative efficiency: Demonstrates exponential time-efficiency in likelihood computation on complex networks compared to MCMC_BiMarkers from PhyloNet.
• Data types: Designed for high-quality DNA sequences and complete genomes using biallelic markers.
• Performance: Shown in simulations to recover simple networks and to achieve higher accuracy and faster likelihood calculations on more complex scenarios than MCMC_BiMarkers.

Scientific Applications:

• Reticulate evolution inference: Inferring horizontal gene transfer, hybridization, and introgression in species histories.
• Genome-scale marker analysis: Analyzing biallelic markers from complete genomes and high-quality DNA sequences.
• Rice evolution case study: Applied to a rice dataset to infer evolutionary scenarios consistent with previous findings and to provide additional insights into rice evolution.

Methodology:

Operates within BEAST 2, integrates Snapp and SpeciesNetwork, uses MCMC operators from SpeciesNetwork and Snapp-specific operators for population sizes and mutation rates, computes likelihoods of biallelic markers under the multispecies network coalescent (MSNC), and performs Bayesian inference via MCMC.