Abstract
Targeted GBS is a recent approach for obtaining an effective characterization for hundreds to thousands of markers. The high throughput of next-generation sequencing technologies, moreover, allows sample multiplexing. The aims of this study were to (i) define a panel of single nucleotide polymorphisms (SNPs) in the cat, (ii) use GBS for profiling 16 cats, and (iii) evaluate the performance with respect to the inference using standard approaches at different coverage thresholds, thereby providing useful information for designing similar experiments. Probes for sequencing 230 variants were designed based on the Felis_catus_8.0. 8.0 genome. The regions comprised anonymous and non-anonymous SNPs. Sixteen cat samples were analysed, some of which had already been genotyped in a large group of loci and one having been whole-genome sequenced in the 99_Lives Cat Genome Sequencing Project. The accuracy of the method was assessed by comparing the GBS results with the genotypes already available. Overall, GBS achieved good performance, with 92-96% correct assignments, depending on the coverage threshold used to define the set of trustable genotypes. Analyses confirmed that (i) the reliability of the inference of each genotype depends on the coverage at that locus and (ii) the fraction of target loci whose genotype can be inferred correctly is a function of the total coverage. GBS proves to be a valid alternative to other methods. Data suggested a depth of less than 11× is required for greater than 95% accuracy. However, sequencing depth must be adapted to the total size of the targets to ensure proper genotype inference.