Understanding the genetic relationship between fertility and feed efficiency traits and its pleiotropic effect in cattle

生育率下降是畜牧业盈利能力下降和动物福利下降的主要关切之一。同样，饲料效率的降低对奶牛群的盈利能力和动物福利也有很大的影响。饲料成本最高可占牛群生产成本的75%。有趣的是，在肉牛群和奶牛群中，FE和繁殖力之间存在相关但不利的遗传相关性。这种不利的遗传相关性的遗传基础仍然鲜为人知。因此，更好地了解育性和FE性状之间的遗传相关性所涉及的生物学过程将有助于开发新的模型和策略来改进同时遗传选择。性状间遗传相关的主要生物学过程是多效性。分解与观察到的性状间遗传相关相关的生物过程可以为阐明感兴趣的表型的发育过程带来巨大的好处，并提高对这些性状的选择效率。标记和表型之间的相互作用可以被调节为因果网络，基因与其产物之间的相互作用，如mRNAs，可以被调节为共表达网络。对网络结构中这些生物信息的分析和解释允许应用图论概念，这有助于提取关于假定的因果变异和功能候选基因优先顺序的有意义的信息。在拟议的创新研究计划中，我的目标是识别与多效性效应相关的功能候选基因和遗传变异，这可能是在奶牛群中观察到的受精率和FE之间不利的遗传相关性的原因。这一目标将通过计算多效性分数来实现，这些分数来自对生育力和FE性状进行的全基因组关联研究的汇总统计。最先进的多OMICS技术将用于从怀孕和未怀孕的荷斯坦奶牛的子宫内膜组织以及高和低FE荷斯坦奶牛的乳样中鉴定差异共表达的基因网络。这些信息将使用生物信息学和统计管道进行整合，基于网络表示学习，以确定多效性效应的功能候选基因的优先顺序。这一功能候选基因的鉴定第一次有可能为荷斯坦牛的受精性和FE的因果变异的鉴定提供一个精细的图谱。因此，有助于减少选择偏差，不利于同时选择，适当地考虑了性状之间的相关性，同时有助于提高预测精度、统计能力和参数估计精度。

Reduced fertility is one of the main concerns regarding decreased profitability and animal welfare in the livestock sector. Similarly, reduced feed efficiency (FE) also has a great impact of the profitability of dairy herds and over the animal welfare. Feed costs can represent up to 75% of the production costs in cattle herds. Interestingly, a relevant, but unfavourable, genetic correlation is observed between FE and fertility in both beef and dairy cattle herds. The genetic basis of this unfavourable genetic correlation is still poorly understood. Therefore, the better understanding of the biological processes associated with genetic correlation between fertility and FE traits would help to develop new models and strategies to improve the simultaneous genetic selection. The main biological process responsible for the genetic correlation between traits is the pleiotropy. The decomposition of the biological processes associated with the observed genetic correlations between traits can bring great advantages for elucidating the development process of the phenotypes of interest, as well as increase the efficiency of selection for these traits. The interactions between markers and phenotypes can be modulate as causality networks as well as the interactions between genes and their production, such as mRNAs, can be modulate as co-expression networks. The analysis and interpretation of this biological information in a network structure allows the application of graph theory concepts which can help to extract meaningful information regarding putative causal variants and prioritization of functional candidate genes. In the proposed innovative research program I aim to identify functional candidate genes and genetic variants associated with the pleiotropic effect that might be the cause of the unfavourable genetic correlation observed between fertility and FE in dairy herds. This objective will be reach through calculation of pleiotropic scores from the summary statistics of genome wide association studies performed for fertility and FE traits. State-of-the-art multi-OMICS technologies will be used to identify differentially co-expressed gene networks from the endometrium tissue of pregnant and non-pregnant Holstein cows, as well as from milk samples from high- and low-FE Holstein cows. This information will be integrated using a bioinformatic and statistical pipeline, based on network representation learning, to prioritize functional candidate genes for the pleiotropic effect. The identification of this functional candidate genes has the potential to allow, for the first time, a fine mapping for the identification of causal variants simultaneously responsible for fertility and FE in Holstein cattle. Consequently, helping to reduce selection bias, unfavourable simultaneous selection, appropriately accounts the correlation between the traits, while helping to increase prediction accuracy, statistical power, and parameter estimation accuracy.