Simulating transgenerational hologenomes under selection with RITHMS

TL;DR

RITHMS addresses the need to simulate transgenerational hologenomic data that integrate microbiota transmission, environment, and selection. It extends MoBPS by incorporating a microbiota compartment, environmental covariates, and genetic modulation of taxa, yielding a phenotype modeled as $y^{(t)} = oldsymbol{b1}^T oldsymbol{G}^{(t)} + oldsymbol{}ty^T oldsymbol{B}^{(t)} + oldsymbol{b epsilon}_y^{(t)}$, and it supports calibrating direct heritability $h^2_d$ and microbiability $b^2$ to reflect hologenomic contributions. Using real base data for initialization, RITHMS simulates multiple generations under diverse breeding schemes, including mixed-objective selection that balances phenotypic gain and microbial diversity. The framework demonstrates how vertical and environmental transmissions shape microbiota structure, diversity, and host phenotypes, and provides tools for exploring partially observed microbiota data. Overall, RITHMS offers a flexible, open-source platform to generate transgenerational hologenomes under selection, enabling evaluation of hologenomic strategies and their practical implications for breeding programs.

Abstract

A holobiont is made up of a host organism together with its microbiota. In the context of animal breeding, the holobiont can be viewed as the single unit upon which selection operates. Therefore, integrating microbiota data into genomic prediction models may be a promising approach to improve predictions of phenotypic and genetic values. Nevertheless, there is a paucity of hologenomic transgenerational data to address this hypothesis, and thus to fill this gap, we propose a new simulation framework. Our approach, an R Implementation of a Transgenerational Hologenomic Model-based Simulator (RITHMS) is an open-source package. It builds upon simulated transgenerational genotypes from the Modular Breeding Program Simulator (MoBPS) package and incorporates distinctive characteristics of the microbiota, notably vertical and horizontal transmission as well as modulation due to the environment and host genetics. In addition, RITHMS can account for a variety of selection strategies and is adaptable to different genetic architectures. We simulated transgenerational hologenomic data using RITHMS under a wide variety of scenarios, varying heritability, microbiability, and microbiota transmissibility. We found that simulated data accurately preserved key characteristics across generations, notably microbial diversity metrics, exhibited the expected behavior in terms of correlation between taxa and of modulation of vertical and horizontal transmission, response to environmental effects and the evolution of phenotypic values depending on selection strategy. Our results support the relevance of our simulation framework and illustrate its possible use for building a selection index balancing genetic gain and microbial diversity and for evaluating the impact of partially observed microbiota data. RITHMS is an advanced, flexible tool for generating transgenerational hologenomes under selection that incorporate the complex interplay between genetics, microbiota and environment.

Figures (6)

• Figure 1: Schematic illustration of transgenerational hologenomic simulations with RITHMS. The base generation is calibrated on user-provided data for sire genotypes $\boldsymbol{G}_{\male}^{(0)}$, dam genotypes $\boldsymbol{G}_{\female}^{(0)}$, and microbiota data from dams $\boldsymbol{M}_{\female}^{(0)}$. Genotypes are simulated using MoBPS pookMoBPSModularBreeding2020. The sources contributing to the variability of taxa abundances of individual $i$ at generation $t$ are as follows: (1) vertical transmission from the individual's mother $\boldsymbol{M}_{d(i)}^{(t-1)}$, for example during delivery; (2) horizontal transmission of the individual-specific ambient microbiota $\boldsymbol{M}_{a(i)}^{(t)}$; (3) the host selective filter, through which the host's genotype $\boldsymbol{G}_i^{(t)}$ facilitates the colonization and establishment of certain microorganisms; and (4) individual-specific environmental effects $\mathbf{E}_i^{(t)}$, such as diet or treatment effects, modulating the microbiota composition. Phenotypes $\boldsymbol{y}_i^{(t)}$ are simulated according to a linear model, where the microbiota has a direct effect and the genome has both direct and microbiota-mediated effects perez-encisoOpportunitiesLimitsCombining2021.
• Figure 2: Key characteristics of microbiota data simulated with RITHMS. (A) Pairwise correlation matrix of taxa abundances. Abundances were simulated assuming all taxa are under genetic control and distributed in five clusters (shown with color bars in the margins). Taxa are sorted based on the cluster they belong to. (B) Density plot of the distribution of taxa heritability for increasing genetic effect sizes ($\sigma_{\beta}\times\sqrt{\text{QTL}_\text{o}}$), shown above each curve. (C) Correlation between offspring $\alpha$-diversity (from G2) and that of its mother (purple), father (orange) or ambient microbiota (green) for increasing values of $\lambda$. Correlations are computed from a population of 500 offsprings and averaged over 10 repetitions. (D) Density plots of the distribution of $\alpha$-diversity values in the base population (G0) and five consecutive generations (G1 to G5), in the absence of selection and environmental filters.
• Figure 3: Simulation of sporadic (top) and sustained (bottom) environmental effects in RITHMS. (A) Multidimensional scaling (MDS) of microbial abundance data (Bray-Curtis distances). Half the individuals at G1 (blue triangles) are subject to a sporadic antibiotic treatment. (B) Density plots of $\alpha$-diversity values before (G0), during (G1) and after (G2 to G3) sporadic antibiotic treatment. (C) Multidimensional scaling (MDS) of microbial abundance data (Bray-Curtis distances). Starting from G1, half the individuals at each generation (blue triangles) are subject to a diet favoring two clusters of taxa.(D) Density plots of $\alpha$-diversity values before (G0) and during (G1 to G3) sustained diet intervention.
• Figure 4: Direct heritability and microbiability of RITHMS simulations under various selection strategies. (A) Observed direct heritability $h_\text{d}^2$ and microbiability $b^2$ (averaged over 50 simulated datasets) in a scenario with random selection and target values $h_\text{d}^2 = b^2 = 0.25$.(B) Mean phenotypic change across five generations, (averaged over 50 simulated datasets, shaded regions correspond to 95% confidence intervals) with $\lambda = 0.1$, according to various values of direct heritability (rows) and microbiability (columns) and different selection strategies: $\textbf{BV}_d^{(t)}$ (direct breeding values, blue line), $\textbf{BV}_m^{(t)}$ (microbiota breeding values, red line), $\textbf{BV}_t^{(t)}$ (total breeding values, purple line), random selection of parents for the next generation (black line).
• Figure 5: Simulation-guided exploration of mixed selection index. Mean phenotype and microbial diversity changes from the base population (G0) to G5 as a function of $\textrm{w}_{\textrm{div}}$. The simulation is repeated 25 times for each value of $\textrm{w}_{\textrm{div}}$. Each simulation is shown as semi-transparent dots whereas square dots correspond to the mean computed over the 25 repetitions.