Quality assessment of boar semen by multivariate analysis of flow cytometric data

Abstract

Flow cytometry (FCM) has become very powerful over the last decades, enabling multi-parametric measurements of up to thousands of cells per second. This generates massive amounts of data on individual cell characteristics that need to be analyzed in an efficient manner from both physiological and chemical points of view. In this study, a methodology of analysis for FCM data was comprehensively studied to assess quality changes in semen extracted from boars. The proposed methodology combines new automated multi-dimensional data normalization, a density-based clustering method for identification of cell populations, and multivariate methods for post-analysis of the identified populations, enabling the exploratory evaluation and prediction/classification of subpopulations within the experimental data set. The performance of the suggested methodology was compared with the performance of an existing automated clustering method.