Skip to main content

Table 9 Comparing t-test results for the three different imputation methods

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

Normalization 2 Missing replaced by = -17.28 Log transformed (100% of identified spots picked up) Normalization 2 Missing replaced by Random sample of minimum from 15 gels in each group Log transformed (91% of identified spots picked up) Normalization 2 Minimum replaced by Random sample of minimum from all 30 gels Log transformed (91% of identified spots picked up)
SSP 1134 SSP1134 SSP 1134
SSP 1509 SSP 1509 SSP 1509
SSP 1733 SSP 1733 SSP 1733
SSP 2309 SSP 2309 SSP 2309
SSP 3219   
SSP 3806 SSP 3806 SSP 3806
SSP 4203 SSP 4203 SSP 4203
SSP 4225 SSP 4225 SSP 4225
SSP 4435 SSP 4435 SSP 4435
SSP 4724 SSP 4724 SSP 4724
SSP 5413   SSP 5413
SSP 6205 SSP 6205 SSP 6205
SSP 6304 SSP 6228 SSP 6228
SSP 6314 SSP 6304 SSP 6304
  SSP 6314 SSP 6314
SSP 6452 SSP 6452 SSP 6452
SSP 7223 SSP 7223 SSP 7223
  SSP 7334 SSP 7334
SSP 7750 SSP 7750 SSP 7750
  1. In column 1 missing values were replaced with the lowest intensity value in experiment; in column 2 values to replace missing intensities were randomly chosen from the 15 lowest intensity values within a treatment group; in values to replace missing intensities were randomly chosen from the 30 lowest intensity values without regard to treatment group.