EpiReSIM: A Resampling Method of Epistatic Model without Marginal Effects Using Under-Determined System of Equations.

Area of Science:

• Genetics and Bioinformatics
• Computational Biology
• Statistical Genetics

Background:

• Epistasis detection methods require robust simulation experiments for validation.
• Existing simulators struggle with epistasis models lacking marginal effects (eNME models).
• Lack of effective eNME model simulators hinders progress in epistasis detection.

Purpose of the Study:

• To develop a novel resampling simulation method, EpiReSIM, for generating eNME models.
• To address the limitations of current simulators in handling eNME models.
• To facilitate the validation and application of epistasis detection methods.

Main Methods:

• EpiReSIM employs two strategies for solving eNME models: prevalence constraints and joint prevalence-heritability constraints.
• The method transforms model computation into solving underdetermined systems of equations.
• Utilizes complete orthogonal decomposition and Newton's method for solving equations, enabling high-order model solutions.
• Generates simulation data via a resampling approach based on the computed eNME model.

Main Results:

• EpiReSIM successfully generates eNME models and simulation data.
• Demonstrates advantages in preserving minor allele frequency biological properties.
• Effectively calculates high-order epistasis models.
• Provides a convenient and effective alternative to existing simulation software.

Conclusions:

• EpiReSIM is a valuable tool for simulating epistasis without marginal effects.
• The method enhances the validation capabilities for epistasis detection algorithms.
• EpiReSIM offers a practical solution for generating biologically relevant simulation data.