Roghaye Tanegonbadi1, Hossein Moradi Shahrbabak *1, Mohammad Moradi shahrbabak1, Hossein Bani saadat2
1- University of Tehran
2- Tarbiat Modares University
2- Tarbiat Modares University
Abstract: (34 Views)
Livestock breeding and selection of superior breeders are considered one of the main foundations for improving livestock production and promoting productivity in the livestock industry. In this regard, careful selection of animals as parents of the next generation is of great importance, with the aim of improving economic traits and increasing genetic progress in livestock populations. Reproductive traits in sheep, especially the ability of ewes to produce more lambs per calving, are among the most important economic indicators that directly affect production efficiency and profitability of breeding systems. For this reason, a detailed understanding of the genetic mechanisms controlling these traits can play a decisive role in the design of modern breeding programs. Machine learning methods, especially nonlinear algorithms such as Gradient Boosting, are able to reveal complex patterns in genomic data that are often hidden from the view of traditional statistical models. Therefore, the aim of this study was to identify genes and single nucleotide polymorphisms (SNPs) affecting the litter size trait in Zandi sheep using the gradient boosting algorithm.
The present study was conducted on data obtained from genotyping 99 Zandi sheep at Tehran Zandi Sheep Breeding and Improvement Station (Pishvay Varamin) using Illumina Ovine SNP 50K chips to identify genes for the litter size trait using the gradient boosting method.
Significant markers including OAR13_15404120, OAR14_49049707, OAR19_32369125, OAR5_71435920, OAR12_73124271 and s65981 were identified by machine learning (gradient boosting) for the trait of litter size in Zandi sheep. The markers identified by machine learning methods were associated with novel genes including CELF2, TTC9B, CCNP, MAP3K10, FRMD4B, GABRG2, HSD11B1, CACNA1B. Gene ontology (GO) analyses showed that these genes are involved in numerous biological processes including gene expression regulation, neural signal transduction, hormonal activities, cell growth, regulation of glucocorticoid metabolism and calcium transport pathways. For example, the HSD11B1 gene is involved in regulating the conversion of cortisone to cortisol and controlling physiological responses related to fertility. The CACNA1B gene is also part of the voltage-gated calcium channels that play a role in hormone secretion and transmission of nerve messages effective in reproductive processes. The FRMD4B gene is also involved in regulating cell adhesion and signaling in uterine epithelial cells, which can affect implantation success and embryo development. Comparison of the results with previous studies showed that some of these genes, such as HSD11B1 and CACNA, have also been previously reported to be associated with reproductive traits and litter size in sheep. However, several novel genes, including TTC9B and MAP3K10, were identified for the first time in the Zandi sheep population. It was noteworthy that many of these genes were not identified by traditional linear methods such as GWAS, while the gradient boosting algorithm was able to reveal nonlinear effects and complex gene interactions. These findings indicate the higher power and accuracy of machine learning methods in analyzing high-dimensional genomic data and multifactorial relationships.
In general, the gradient boosting method, with its ability to model nonlinear relationships and identify interactions between SNPs, provides a deeper view of the genetic structure of traits and can be a valuable complement to conventional statistical models such as GWAS. The identification of candidate genes in this study provided new insight into the biological pathways involved in reproductive traits and can be used as a basis for future functional studies and more precise genomic selection programs. In addition, the results of Gene Set Enrichment analysis for the identified genes showed that most of them are active in biological pathways related to hormonal responses, follicular growth, and energy metabolism regulation. Overall, the findings of this study can help clarify the genetic mechanisms controlling reproductive traits in Zandi sheep and pave the way for designing more efficient breeding strategies to increase productivity and improve economic traits in the country's livestock populations.
The present study was conducted on data obtained from genotyping 99 Zandi sheep at Tehran Zandi Sheep Breeding and Improvement Station (Pishvay Varamin) using Illumina Ovine SNP 50K chips to identify genes for the litter size trait using the gradient boosting method.
Significant markers including OAR13_15404120, OAR14_49049707, OAR19_32369125, OAR5_71435920, OAR12_73124271 and s65981 were identified by machine learning (gradient boosting) for the trait of litter size in Zandi sheep. The markers identified by machine learning methods were associated with novel genes including CELF2, TTC9B, CCNP, MAP3K10, FRMD4B, GABRG2, HSD11B1, CACNA1B. Gene ontology (GO) analyses showed that these genes are involved in numerous biological processes including gene expression regulation, neural signal transduction, hormonal activities, cell growth, regulation of glucocorticoid metabolism and calcium transport pathways. For example, the HSD11B1 gene is involved in regulating the conversion of cortisone to cortisol and controlling physiological responses related to fertility. The CACNA1B gene is also part of the voltage-gated calcium channels that play a role in hormone secretion and transmission of nerve messages effective in reproductive processes. The FRMD4B gene is also involved in regulating cell adhesion and signaling in uterine epithelial cells, which can affect implantation success and embryo development. Comparison of the results with previous studies showed that some of these genes, such as HSD11B1 and CACNA, have also been previously reported to be associated with reproductive traits and litter size in sheep. However, several novel genes, including TTC9B and MAP3K10, were identified for the first time in the Zandi sheep population. It was noteworthy that many of these genes were not identified by traditional linear methods such as GWAS, while the gradient boosting algorithm was able to reveal nonlinear effects and complex gene interactions. These findings indicate the higher power and accuracy of machine learning methods in analyzing high-dimensional genomic data and multifactorial relationships.
In general, the gradient boosting method, with its ability to model nonlinear relationships and identify interactions between SNPs, provides a deeper view of the genetic structure of traits and can be a valuable complement to conventional statistical models such as GWAS. The identification of candidate genes in this study provided new insight into the biological pathways involved in reproductive traits and can be used as a basis for future functional studies and more precise genomic selection programs. In addition, the results of Gene Set Enrichment analysis for the identified genes showed that most of them are active in biological pathways related to hormonal responses, follicular growth, and energy metabolism regulation. Overall, the findings of this study can help clarify the genetic mechanisms controlling reproductive traits in Zandi sheep and pave the way for designing more efficient breeding strategies to increase productivity and improve economic traits in the country's livestock populations.
Type of Study: Research |
Subject:
ژنتیک و اصلاح نژاد دام
Received: 2025/11/19 | Accepted: 2026/06/29
Received: 2025/11/19 | Accepted: 2026/06/29
Send email to the article author
| Rights and permissions | |
 |
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License. |
