Identification of structural variants and selection signatures in cattle

Given the impacts caused by climate change upon livestock production, it is important to characterize the cattle genome to unravel the genetic mechanisms underlying phenotypic variation that were influenced by the environment and shaped by natural selection that allowed them to thrive in distinct ecosystems. Therefore, the objective of this study is to describe the main effects of adaptation and selection in indicine and locally adapted taurine cattle breeds through the identification of structural variants and signatures of selection using genotypic and whole-genome re-sequencing data. In chapter 2, imputed genotypes (n=735,044 markers) were used to assess genome-wide autozygosity based on runs of homozygosity (ROH) in 9,386 Nellore animals and its lineages using the Plink software. Overall, inbreeding coefficients based on ROH (FROH) were not high, with values close to 2%. Autozygosity islands were evident across the genome, and their genomic location did not largely differ within lineages. Enriched terms (p<0.01) within the autozygosity islands suggested a strong selection for immune response-related traits and might explain the greater adaptability of the indicine cattle in harsh environments. Chapter 3 aimed to assess genome-wide autozygosity to explore ROH hotspot regions which could better characterize the different biological types (productive or adaptive) within the composite Montana Tropical® beef cattle. Montana animals (n=1,436) were genotyped with the GGP-LD BeadChip (n=30,105 markers), and ROH were identified in every individual using the Plink software. The number of autozygosity islands did not differ considerably between biological types, and no significant enriched term (p<0.05) was found to be shared between them. Enriched terms associated with the immune response and homeostasis were described for the adaptive biological type, while those linked to the immune system as well as with reproductive and productive functions we identified for the productive biological type. In chapter 4, four statistical methods were implemented to detect genomic regions under selective pressure using whole-genome re-sequencing data from Gir (GIR, n=13), Caracu Caldeano (CAR, n=12), Crioulo Lageano (CRL, n=12), and Pantaneiro (PAN, n=12) cattle breeds. Within-population (CLR and iHS) and cross-population statistics (FST and XPEHH) were combined separately in a single score using the de-correlated composite of multiple signals (DCMS) method, and putative sweep regions were revealed by assessing the top 1% of the empirical distribution generated by each DCMS statistic. The signatures of selection identified herein provided a comprehensive set of putative candidate genes together with QTLs disclosing cattle production traits and adaptation to the challenging environment in which these breeds have been exposed. In chapter 5, the read depth-based method implemented in CNVnator was used for copy number variants (CNV) calling on resequenced data (~14.07 X) from CAR (n=12), CRL (n=12), and PAN (n=12) cattle breeds. CNV regions (CNVRs) were identified by overlapping individual CNVs within each breed. The functional annotation of the CNVRs revealed variants with high consequence on protein sequence harboring relevant genes with functions strongly linked to environmental resilience (i.e., BOLA-DQB, BOLA-DQA5, CD1A, β-defensins, PRG3, and ULBP21). Enrichment analysis based on the gene list retrieved from the viii CNVRs also disclosed over-represented terms (p<0.01) greatly associated with immunity and cattle resistance to harsh environments. Our findings improve the knowledge about the genome biology of such cattle breeds and provide candidate genes and genomic regions encompassing relevant traits as well as useful information for future conservation, association, or selection approaches. (AU)