These differences accounted for 49% of total variation in ubiquitin-induced fluorescence in this
trial. However, fluorescence intensity of the ubiquitinated sperm population (M3-median) did not
differ among treatments (P = 0.92). As expected, the percentage of presumed normal
spermatozoa lacking ubiquitin-induced fluorescence (% M2) was significantly higher (P < 0.001)
in nanopurified samples than in controls and accounted for 44% of total variation in ubiquitin
negative immunofluorescence in the study (Fig. 3 b).