Skip to main content

Table 10 Comparing variances of spot intensities when using multiple values to impute versus a single value imputation.

From: The case for well-conducted experiments to validate statistical protocols for 2D gels: different pre-processing = different lists of significant proteins

  Multiple Imputation Single Imputation  
  Treatment Control Treatment Control  
SSP 1509 0.767 0.663 7.05 8.39 9 missing control
  Range = (3.16) Range = (2.63) Range = (7.34) Range = (6.56) 3 missing treatment
SSP 1733 0.356 1.4249 0.356 1.4249 0 missing control
  Range = (2.23) Range = (3.74) Range = (2.23) Range = (3.74) 0 missing treatment
SSP 6314 1.2055 1.178 3.41 1.17 0 missing control
  Range = (3.94) Range = (3.51) Range = (7.15) Range = (3.51) 14 missing treatment
SSP 6452 0.547 0.564 0.547 0 15 missing control
  Range = (2.79) Range = (2.14) Range = (2.79) Range = (0) 0 missing in treatment
  1. The variances for spots with missing values are either under estimated or over estimated with single imputation values. The ranges and variance values of intensities are closer to that of SSP 1733 (spot with no missing intensities) in the case of multiple value imputation.