Multi-omics analysis

Evaluating the contribution of microbiome information to genetic and genomic evaluations for methane emissions in growing cattle

---

Abstract

The integration of microbiome information into genetic evaluation frameworks offers potential to improve prediction accuracy for complex traits influenced by host-microbe interactions. This study investigated the contribution of rumen microbiome profiles to methane genetic and genomic evaluations in beef cattle using pedigree-based (BLUP), genomic (GBLUP), and single-step genomic (ssGBLUP) models. Phenotypic, genomic, and microbial data were available for 810 growing beef cattle, fed a high-energy finishing diet, at the Irish Cattle Breeding Federation Progeny Test Centre, Kildare, Ireland. Methane emissions were measured using the GreenFeed system (C-Lock Inc., Rapid City, SD), with average methane production calculated per animal across the 21-100d test period. Rumen digesta samples were collected at the end of methane measurement, and microbial DNA was extracted and sequenced targeting the V4 region of 16S rRNA gene. After quality control, 264 amplicon sequence variants (ASVs) were retained, and a microbiome relationship matrix (MRM) was computed as a variance-covariance matrix of standardised ASVs. Host genotypes were generated using a medium density 50K SNP chip and used to compute a genomic relationship matrix. Six animal model scenarios were evaluated: pedigree (BLUP), genomic (GBLUP), and unified pedigree and genomic (ssGBLUP), each with and without inclusion of the MRM. Variance components were estimated using univariate pedigree and genomic animal models. Pedigree was traced back to founder animals. Predictive ability of methane estimated breeding values (EBVs) from each model scenario were assessed by fourfold cross-validation, where the genetic merit of the validation animals was predicted from relatives' performance in each fold. Inclusion of the MRM altered variance component partitioning across scenarios. In the GBLUP model, the heritability of methane decreased from 0.59±0.12 to 0.37±0.11, and residual variance reduced from 41% to 37% of phenotypic variance, with the inclusion of microbial information; the microbiome explained 26±4% of phenotypic variance. A similar pattern occurred in the single-step genomic framework, where h² declined from 0.56±0.12 to 0.35±0.10 following MRM inclusion. Cross-validation results indicated comparable (G)EBV predictive ability among model scenarios. Correlations between adjusted phenotypes and predicted methane (G)EBVs of validation animals ranged from 0.17±0.02 (BLUP) to 0.21±0.02 (ssGBLUP), while dispersion (≈1.0) and level bias (≈0) remained near optimal across models. Inclusion of microbiome information did not significantly (P >0.05) alter slope, bias, or dispersion statistics between models. Overall, incorporating microbiome relationships into pedigree and genomic evaluations redistributed variance components, with part of the previously additive genetic variance captured by microbial similarity, suggesting meaningful host-microbiome interactions. Gains in predictive ability were low to moderate, yet the absence of performance loss indicates microbiome inclusion is feasible and potentially informative. Further research is warranted to assess the stability of microbiome effects across different management systems, particularly under grazing conditions.

Keywords: 2026

How to Cite:

Kelly, D., Smith, P., Kirwan, S., Waters, S., Conroy, S. & Evans, R., (2026) “Evaluating the contribution of microbiome information to genetic and genomic evaluations for methane emissions in growing cattle”, World Congress on Genetics Applied to Livestock Production Digital Archive 2026(1): 2286655. doi: https://doi.org/10.31274/wcgalp.24082

Rights: 1