A leap forward: exploring the advantages of single-step genome evaluation in Italian Mediterranean buffalo

Stefano Biffani, Mayra Gómez Carpio, Roberta Cimmino, Dario Rossi, G. Zullo, Riccardo Negrini, Alberto Cesarani, Giuseppe Campanile, Gianluca NegliaRevista 2023 · DOI

Abstract

The implementation of the Single-Step Genomic Best Linear Unbiased Predictor (ssGBLUP) represents a significant advancement in the genetic evaluation of the Italian Mediterranean Buffalo (IMB). By integrating genomic information with traditional pedigree data, this study demonstrates a marked increase in the accuracy of breeding value (BV) estimations for key production and morphological traits. Results indicate that ssGBLUP outperforms traditional methods, particularly for economically vital traits like mozzarella yield and milk protein content, offering a more robust framework for future genetic progress.

Introduction: The Challenge of Genetic Selection

In livestock breeding, identifying animals with superior genetic merit—expressed as Estimated Breeding Values (EBVs)—is crucial for herd improvement. Traditionally, the Italian Mediterranean Buffalo industry relied on the Best Linear Unbiased Predictor (BLUP) method. While effective, BLUP depends on pedigree records, which cannot account for the stochastic nature of gene inheritance, known as Mendelian sampling.

With the advent of genomic sequencing, DNA-based information offers a clearer picture of an animal's true genetic potential. However, genotyping an entire population is often cost-prohibitive. The challenge lies in effectively combining genomic data from a subset of animals with the extensive phenotypic and pedigree records of the broader population to improve selection accuracy.

Methodology: The Single-Step Approach

This study evaluated the feasibility of the ssGBLUP method. Unlike multi-step methods that analyze genomic and pedigree data separately, ssGBLUP utilizes a single, integrated relationship matrix, known as the H matrix. This matrix is created by mathematically blending the traditional numerator relationship matrix (A), based on pedigrees, with the genomic relationship matrix (G), based on DNA markers.

To validate the predictive power of this model, researchers employed the LR (Linear Regression) method. This involved creating "partial" datasets by masking recent records (using cut-off years 2012, 2015, and 2017) and comparing the resulting predictions against the "whole" dataset. This approach measures the accuracy, bias, and dispersion of the breeding values over time.

Data Overview

The study leveraged one of the most comprehensive longitudinal datasets for this breed:

  • Production Traits: 743,904 lactation records from 276,451 buffalo cows (1984–2019), including milk volume, fat/protein percentages, and mozzarella yield (MY).
  • Morphological Traits: 91,966 records (2004–2022) focusing on feet, legs, and the mammary system.
  • Genomic Data: DNA profiles for 2,017 cows and 133 bulls.

Key Results and Discussion

The transition to ssGBLUP yielded substantial improvements across all evaluated metrics compared to traditional pedigree-based BLUP:

1. Significant Accuracy Gains

The most striking results were observed in milk composition and mozzarella yield. For these traits, the accuracy of breeding value predictions increased by more than 20% when using ssGBLUP. The correlation between predicted and actual performance for production traits rose from a baseline of 6% in traditional models to 49% in the genomic model.

2. The "Halo Effect" for Non-Genotyped Animals

A critical advantage of the single-step method is its impact on non-genotyped animals. Because the H matrix links genotyped individuals to their non-genotyped relatives, the accuracy of EBVs improved for the entire population, including females without DNA records. This ensures that genomic investments benefit the management of the whole herd.

3. Reduced Bias and Improved Reliability

The LR statistics showed a general reduction in prediction bias and dispersion. This indicates that ssGBLUP provides more stable and reliable estimates of genetic merit, reducing the risk of selecting animals based on over- or under-estimated potential.

Conclusion

The implementation of ssGBLUP in the Italian Mediterranean Buffalo breeding program is not only feasible but essential for accelerating genetic gain. By providing more accurate predictions for high-value traits like mozzarella yield, this genomic approach offers significant economic benefits to the industry. Future efforts should focus on expanding the genotyped reference population and applying this methodology to harder-to-measure traits such as fertility and longevity.

Summary updated

Abstract

Single-step genomic best linear unbiased predictor (ssGBLUP) is a method for jointly estimating breeding values (BV) for genotyped and non-genotyped animals. Genomic information in the Italian Mediterranean Buffalo (IMB) is now available and its inclusion in the genetic evaluation system could increase both accuracy and genetic progress of the traits of interest of the breed. The aim of this study was to test the feasibility of ssGBLUP and to present the first results of the implementation of a genomic evaluation for production and type traits in the IMB. Phenotypic information on production (270-day milk, mozzarella yield (MY), protein and fat kg and %, respectively) and morphology: feet and legs (FL) and mammary system (MS) were used for this study. Production records included 743,904 lactations from 276,451 buffalo cows born from 1984 to 2019. Morphological traits were from 91,966 buffalo cows from 2004 to 2022. Regarding the genotypes, a total of 2,017 buffalo cows and 133 bulls were used. Data were analysed fitting two multitrait animal models, a 6-trait model for production data and a 2-trait model for morphology data. According to the relationship matrix used, two models were fitted: (i) the BLUP with the numerator relationship matrix (A); (ii) the ssGBLUP where A and the genomic relationship matrix (G) are blended into H. BV were estimated with BLUP and ssGBLUP models. Three different scenarios were used, according to the cut-off year used to create the partial datasets, namely 2012, 2015 and 2017. In each scenario, correlation, accuracy, dispersion, and bias statistics were calculated (LR method). Both bulls (N=49) and cows (N=1288) were used for validations. On average, correlation between EBVs from partial and whole datasets estimated with BLUP and ssGBLUP increased from 6 to 49% and from 14 to 17% for production and type traits, respectively. Among the traits analysed, the ones most affected by the change were protein/fat content, MY, and AM. The accuracy increase for these traits was above 20% when using the ssGBLUP. All LR statistics improved also for non-genotyped females. These results showed that implementing ssGBLUP in the breeding program can generate more accurate predictions for important traits in dairy IMB than traditional BLUP.
BibTeX
@inproceedings{biffani2023leap,
  author = {Stefano Biffani and Mayra Gómez Carpio and Roberta Cimmino and Dario Rossi and G. Zullo and Riccardo Negrini and Alberto Cesarani and Giuseppe Campanile and Gianluca Neglia},
  title = {{A leap forward: exploring the advantages of single-step genome evaluation in Italian Mediterranean buffalo}},
  booktitle = {Revista},
  year = {2023},
  doi = {10.52973/rcfcv-wbc129},
}
Figure