Background: Transforming large amounts of genomic data into valuable knowledge for predicting complex traits has been an important challenge for animal and plant breeders. Prediction of complex traits has not escaped the current excitement on machine-learning, including interest in deep learning algorithms such as multilayer perceptrons (MLP) and convolutional neural networks (CNN). The aim of this study was to compare the predictive performance of two deep learning methods (MLP and CNN), two ensemble learning methods [random forests (RF) and gradient boosting (GB)], and two parametric methods [genomic best linear unbiased prediction (GBLUP) and Bayes B] using real and simulated datasets. Methods: The real dataset consisted of 11,790 Holstein bulls with sire conception rate (SCR) records and genotyped for 58k single nucleotide polymorphisms (SNPs). To support the evaluation of deep learning methods, various simulation studies were conducted using the observed genotype data as template, assuming a heritability of 0.30 with either additive or non-additive gene effects, and two different numbers of quantitative trait nucleotides (100 and 1000). Results: In the bull dataset, the best predictive correlation was obtained with GB (0.36), followed by Bayes B (0.34), GBLUP (0.33), RF (0.32), CNN (0.29) and MLP (0.26). The same trend was observed when using mean squared error of prediction. The simulation indicated that when gene action was purely additive, parametric methods outperformed other methods. When the gene action was a combination of additive, dominance and of two-locus epistasis, the best predictive ability was obtained with gradient boosting, and the superiority of deep learning over the parametric methods depended on the number of loci controlling the trait and on sample size. In fact, with a large dataset including 80k individuals, the predictive performance of deep learning methods was similar or slightly better than that of parametric methods for traits with non-additive gene action. Conclusions: For prediction of traits with non-additive gene action, gradient boosting was a robust method. Deep learning approaches were not better for genomic prediction unless non-additive variance was sizable.

Fat-tail content of sheep breeds is varied and the molecular mechanisms regulating fat-tail development have not been well characterized. Aiming at better identifying the important candidate genes and their functional pathways contributing to fat deposition in the tail, a comparative transcriptome analysis was performed between fat- (Lori-Bakhtiari) and thin-tailed (Zel) Iranian sheep breeds using RNA-seq. The experiment was conducted on six male lambs (three lambs per each breed) at seven months of age. Four different combinations of aligners and statistical methods including Hisat2 + edgeR, Hisat2 + DESeq2, STAR + edgeR and STAR + DESeq2 were used to identify the differentially expressed genes (DEGs). The DEGs were selected for functional enrichment analysis and protein-protein interaction (PPI) network construction. Module analysis was also conducted to mine the functional sub-networks from the PPI network. In total, 264 genes including 80 up- and 184 down-regulated genes were identified as DEGs. The RNA-Seq results were validated by Q-RT-PCR. Functional analysis of DEGs and the module analysis of PPI network demonstrated that in addition to pathways affecting lipid metabolism, a series of enriched functional terms related to “response to interleukin”, “MAPK signaling pathways”, “Wnt signaling pathway”, “ECM-receptor interaction”, “regulation of actin cytoskeleton”, and “response to cAMP” might contribute to the deposition of fat in tails of sheep. Overall results using RNA-Seq analysis characterized important candidate genes involved in the fatty acid metabolism and regulation of fat deposition, suggesting novel insights into molecular aspects of fat-tail metabolism in sheep. Selected DEGs should be further investigated as potential markers associated with the fat-tail development in sheep breeds.

Persian plateau (including Iran) is considered as one of the primary centers of origin of walnut. Sampling walnut trees originating from this arena and exploiting the capabilities of next-generation sequencing (NGS) can provide new insights into the degree of genetic variation across the walnut genome. The present study aimed to explore the population structure and genomic variation of an Iranian collection of Persian walnut (Juglans regiaL.) and identify loci underlying the variation in nut and kernel related traits using the new AxiomJ. regia700K SNP genotyping array. We genotyped a diversity panel including 95 walnut genotypes from eight Iranian provinces with a variety of climate zones. A majority of the SNPs (323,273, 53.03%) fell into the “Poly High Resolution” class of polymorphisms, which includes the highest quality variants. Genetic structure assessment, using several approaches, divided the Iranian walnut panel into four principal clusters, reflecting their geographic partitioning. We observed high genetic variation across all of the populations (H_O = 0.34 and H_E = 0.38). The overall level of genetic differentiation among populations was moderate (F_ST = 0.07). However, the Semnan population showed high divergence from the other Iranian populations (on average F_ST = 0.12), most likely due to its geographical isolation. Based on parentage analysis, the level of relatedness was very low among the Iranian walnuts examined, reflecting the geographical distance between the Iranian provinces considered in our study. Finally, we performed a genome-wide association study (GWAS), identifying 55 SNPs significantly associated with nut and kernel-related traits. In conclusion, by applying the novel AxiomJ. regia700K SNP array we uncovered new unexplored genetic diversity and identified significant marker-trait associations for nut-related traits in Persian walnut that will be useful for future breeding programs in Iran and other countries.

The genomic prediction of unobserved genetic values or future phenotypes for complex traits has revolutionized agriculture and human medicine. Fertility traits are undoubtedly complex traits of great economic importance to the dairy industry. Although genomic prediction for improved cow fertility has received much attention, bull fertility largely has been ignored. The first aim of this study was to investigate the feasibility of genomic prediction of sire conception rate (SCR) in US Holstein dairy cattle. Standard genomic prediction often ignores any available information about functional features of the genome, although it is believed that such information can yield more accurate and more persistent predictions. Hence, the second objective was to incorporate prior biological information into predictive models and evaluate their performance. The analyses included the use of kernel-based models fitting either all single nucleotide polymorphisms (SNP; 55K) or only markers with presumed functional roles, such as SNP linked to Gene Ontology or Medical Subject Heading terms related to male fertility, or SNP significantly associated with SCR. Both single- and multikernel models were evaluated using linear and Gaussian kernels. Predictive ability was evaluated in 5-fold cross-validation. The entire set of SNP exhibited predictive correlations around 0.35. Neither Gene Ontology nor Medical Subject Heading gene sets achieved predictive abilities higher than their counterparts using random sets of SNP. Notably, kernel models fitting significant SNP achieved the best performance with increases in accuracy up to 5% compared with the standard whole-genome approach. Models fitting Gaussian kernels outperformed their counterparts fitting linear kernels irrespective of the set of SNP. Overall, our findings suggest that genomic prediction of bull fertility is feasible in dairy cattle. This provides potential for accurate genome-guided decisions, such as early culling of bull calves with low SCR predictions. In addition, exploiting nonlinear effects through the use of Gaussian kernels together with the incorporation of relevant markers seems to be a promising alternative to the standard approach. The inclusion of gene set results into prediction models deserves further research.