# Modeling genotype-by-environment interactions across climatic conditions reveals environment-specific genomic regions and candidate genes underlying feed efficiency traits in tropical beef cattle

## TL;DR

This study identifies genes and genomic regions linked to feed efficiency in tropical beef cattle under different climate conditions, revealing how these traits respond to heat stress.

## Contribution

The study introduces a novel approach to modeling genotype-by-environment interactions in tropical beef cattle to uncover environment-specific genomic regions and candidate genes for feed efficiency.

## Key findings

- Genomic regions and candidate genes associated with feed efficiency traits in Nellore cattle were identified under varying thermal stress conditions.
- Gene networks for residual feed intake reconfigured with increasing heat stress, while dry matter intake networks remained stable in cooler conditions.
- Environment-specific and shared genes were linked to biological processes like metabolism, stress response, and growth regulation.

## Abstract

Heat stress represents a major limitation for livestock production systems, negatively affecting feed efficiency, animal health and welfare, and overall performance. In this context, the objective of this study was to identify genomic regions, candidate genes, biological pathways, and functional networks associated with dry matter intake (DMI) and residual feed intake (RFI) in Nellore cattle exposed to varying levels of thermal stress. The dataset comprised records from 22,838 animals, with genotypes available for 18,567 individuals. The data were collected during 296 feed efficiency trials between 2011 and 2023 across 21 Brazilian farms. Genome-wide association studies (GWAS) were performed using the single-step GBLUP (ssGBLUP) approach to account for genotype-by-environment (G×E) interactions in Nellore cattle. Environmental variation was modeled using the temperature-humidity index (THI) as the environmental gradient, with analyses stratified across three environmental gradients (EG): low (THI = 66), medium (THI = 74), and high (THI = 81). Fifty-one SNPs were significantly associated with RFI, including 27 shared across all three EGs, 10 exclusive to the low EG, one to the high EG, and 13 shared between the moderate and high EGs. These associations were mapped to 44 candidate genes, with 19 genes commonly identified across all EGs, including key candidates such as PIPOX, GTF2F2, KCTD4, MYO18A, and NFIA. For DMI, 136 significant SNPs were identified: 12 and 39 exclusive to the low and moderate EGs, respectively; 28 shared across all EGs; and 57 shared between the moderate and high EGs. These variants were linked to 58 candidate genes, of which 19 were common to all EGs, including NCAPG, LCORL, FAM13A, HERC3, CCND1, and FGF19. Gene network analyses revealed a clear reconfiguration of interaction structures across thermal gradients, particularly for RFI, where gene connectivity declined with increasing THI levels. For DMI, gene networks remained highly integrated, especially in the lowest THI level. Functional annotation highlighted both conserved and environment-specific regulatory architectures, involving key biological processes such as growth regulation, lipid and protein metabolism, intracellular signaling, stress response, and neuroendocrine control. These findings uncover the environmental sensitivity of RFI and DMI, highlight the complex and dynamic genomic basis of these traits under varying climatic conditions, and support the identification of candidate genes for genomic selection programs aiming to enhance climatic resilience in tropical beef cattle.

The online version contains supplementary material available at 10.1038/s41598-025-33952-1.