GTmix

GTmix infers population admixture networks from local gene genealogies reconstructed from haplotypes using coalescent-based maximum likelihood methods.

Key Features:

• Local Gene Genealogies: Utilizes local gene genealogies inferred from haplotypes that encapsulate evolutionary history and linkage disequilibrium (LD) information.
• Coalescent-Based Inference: Performs maximum likelihood inference under the multispecies coalescent (MSC) model to deduce admixture networks.
• Likelihood Computation Optimization: Incorporates advanced techniques to expedite likelihood computations on the MSC model and to optimize network search algorithms.
• Improved Accuracy with Smaller Datasets: Simulation evaluations indicate GTmix can infer more accurate admixture networks using substantially smaller datasets compared to existing methods.
• Population-Scale Applicability: Designed to handle contemporary population genetic datasets for reconstructing complex demographic histories involving admixture events.

Scientific Applications:

• Admixture Network Reconstruction: Reconstructs population demographic histories and admixture events among populations.
• Genetic Structure and History Inference: Leverages detailed genealogical and LD information from haplotypes to resolve genetic structure and historical relationships beyond single-locus methods.

Methodology:

GTmix infers local gene genealogies from haplotypes and applies coalescent-based maximum likelihood inference under the multispecies coalescent (MSC), with optimized likelihood computations and network search algorithms, and performance evaluated by simulations.