Abstract
Including microbiome information in breeding schemes may be helpful to improve the selection response of livestock populations. However, the complexity of the microbiome makes modeling across species and traits difficult. The estimation of the microbiability and the identification of the microbial species are highly dependent on the methodology used. Indeed, it is complicated to decide which is the best method because we fail to know the true underlying scenario. This study proposes an R package named HoloSimR for simulating the coevolution of the genome and the microbiota under a selection process. HoloSimR allows the user to explore the effect of the microbiota on the phenotypic response to selection and the effects of the environment, host genetics, and symbiosis between microbial species on the composition of the microbiota. HoloSimR demonstrated strong computational performance even under complex simulation settings. To assess its efficiency, a divergent selection process was simulated over ten generations across three different scenarios. These scenarios integrated genetic, microbiota, and hologenome-based phenotypic models, including real data-based microbiota structure and heritability. The simulation of those scenarios took 68.42 min on a standard laptop with 16 GB of RAM. Despite the complexity, the package effectively reproduced real microbiota distributions, heritability structures, and interspecies correlations, confirming its scalability and robustness. HoloSimR provides a valuable research platform, allowing researchers to test hypotheses and develop new approaches in a controlled in silico environment before applying them to real-world breeding programmes. This ultimately advances our understanding of host-microbiota interactions in the context of animal breeding.