Assignment tests using genetic information to establish population membership of individuals, provide the most direct methods to determine the population of origin of unknown individuals. The identification of the breed or population of origin of individuals potentially offers unbiased tools in livestock and is useful in a variety of biological contexts. The aim of this study was to identify the minimum number of informative SNPs from highthroughput genotyping data in Sicilian dairy sheep breeds, and to investigate their usefulness for breed assignment purposes. Individuals of Valle del Belice (48), Comisana (48) and Pinzirita (53) sheep breeds were genotyped using Illumina OvineSNP50K Genotyping BeadChip. We used as input data the normalized log R Ratios, instead of the genotypic data, in order to work with numerical variables, analytically more treatable. A mixed strategy, using Regularized K-means, Penalized Multinomial Regression and Linear Discriminant Analysis, was applied. The method of Regularized K-means reduced the number of useful SNPs at 1,738. On a reduced set of SNPs, the Penalized Multinomial Regression has been performed, with a selection of 100 SNPs. Further, a classical Linear Discriminant Analysis has been applied to reduce again the number of SNPs, obtaining a set of 40 markers. The study demonstrated that the combination of these methods was able to efficiently distinguish the individuals of the breeds involved. In fact, using a leave-one-out cross validation technique, this reduced set of SNPs classified correctly 100% of individuals to their breeds of origin. Regularized Kmeans and Penalized Multinomial Regression perform selection through an L1 penalization, while the third step, Linear Discriminant Analysis, provides a further selection with classical techniques. The results laid the basis for the potential use ofthese selected SNPs to generate a panel that may be used for assignment of parentage or within an industrial setting for tracing the origin of animal products derived from the three breedsinvolved in the study.
|Numero di pagine||1|
|Stato di pubblicazione||Published - 2015|