Nicolò Pietro Paolo Macciotta

Aim of this study was to analyse the genetic background of milk Fourier transform infrared (FTIR) spectra in dairy sheep. Individual milk FTIR spectra, with 1060 wavenumbers each, were available for 793 adult Sarda breed ewes genotyped at 45,813 SNP. The absorbance values of each wavenumber was analysed using a linear mixed model that included dim class, parity and lambing month as fixed effects and flock‐test date and animal as random effects. The model was applied to estimate variance components and heritability and to perform a genome‐wide association study for each wavenumber. Average h 2 of wavenumbers absorbance was 0.13 ± 0.08, with the largest values observed in the regions associated with the characteristic bonds of carbonylic and methylenic groups of milk fat (h 2 = 0.57 at 1724–1728 cm −1 ; and h 2 = 0.34 at 2811–2834 cm −1 , respectively). The absorbance values of wavenumbers were moderately correlated with the estimated heritabilities. After the Bonferroni correction, a total of nine markers were found to be significantly associated with 32 different wavenumbers. Of particular interest was the SNP s63269.1, mapped on chromosome 2, that was found to be associated with 27 wavenumbers. Genes previously found to be related to traits of interest (e.g. disease resistance, milk yield and quality, cheese firmness) are located close to the significant SNP. As expected, the heritability estimated for the absorbance of each wavenumbers seems to be associated with the related milk components.