BESSiE: a software for linear model BLUP and Bayesian MCMC analysis of large-scale genomic data

Area of Science:

• Quantitative Genetics
• Bioinformatics
• Statistical Genomics

Background:

• Genomic marker data necessitates advanced Bayesian algorithms for marker effect estimation.
• Existing software often lacks comprehensive algorithm implementation, scalability, or efficient handling of large genomic datasets.
• Script-based environments like R may struggle with computational demands of large-scale genomic analyses.

Purpose of the Study:

• To introduce BESSiE, a novel software package for genomic data analysis.
• To provide a unified platform for implementing and comparing various Bayesian and BLUP algorithms.
• To address limitations in current software for handling large-scale genomic marker data and complex models.

Main Methods:

• BESSiE implements Best Linear Unbiased Prediction (BLUP) and Bayesian Markov Chain Monte Carlo (MCMC) analyses for linear mixed models.
• Supports multivariate, repeated, and missing observations, alongside continuous and categorical factors.
• Includes algorithms such as genomic BLUP, SNP-BLUP, BayesA, BayesB, BayesCπ, and BayesR for marker effect estimation.

Main Results:

• BESSiE accommodates large-scale genomic marker data and complex models without hard-coded limitations.
• The software enables comparison of diverse Bayesian and BLUP algorithms through simple parameter file modifications.
• It is designed for parameter file-driven, command-line operation in Linux environments.

Conclusions:

• BESSiE provides a versatile solution for researchers analyzing large genomic datasets.
• Facilitates direct comparison of multiple marker effect estimation algorithms within a single software.
• Offers a scalable and efficient alternative to existing tools for complex genomic mixed models.