A generalized framework for AMOVA with multiple hierarchies and ploidies.

Area of Science:

• Population Genetics
• Molecular Ecology
• Bioinformatics

Background:

• The analysis of molecular variance (AMOVA) is a cornerstone statistical method in population genetics and molecular ecology.
• Classic AMOVA frameworks are limited to haploid/diploid data and restricted hierarchical levels (2-4).
• Natural populations often exhibit complex hierarchical structures and varying ploidy levels, even within species or individuals, exceeding the capabilities of traditional AMOVA.

Purpose of the Study:

• To generalize the AMOVA framework to accommodate any number of hierarchical levels and any ploidy level.
• To develop robust statistical methods for analyzing multilocus genotypic and allelic phenotypic data with unknown allele dosage.
• To provide a user-friendly software package for applying these advanced AMOVA methods.

Main Methods:

• Generalized the AMOVA framework to support arbitrary ploidy levels and hierarchical structures.
• Developed four distinct statistical methods to analyze diverse genetic data types, including those with unknown allele dosage.
• Validated the framework and methods using simulated datasets and an empirical dataset.

Main Results:

• The generalized AMOVA framework successfully accommodates complex population structures and polyploidy.
• The developed methods accurately analyze multilocus genotypic and allelic phenotypic data.
• Performance evaluation using simulated and empirical data confirmed the framework's efficacy.

Conclusions:

• The generalized AMOVA framework significantly expands the applicability of population genetic analyses.
• The new methods and the 'polygene' software package offer powerful tools for molecular ecology research.
• This work addresses limitations in existing AMOVA methods, enabling more comprehensive studies of natural populations.