Skip to main content

Table 2 Highly significant (p ≤ 0.001) probe sets lost at different filtering thresholdsa.

From: Effects of filtering by Present call on analysis of microarray experiments

  Unfiltered (total) FP>0
sig254
RMA4.25
FP25
sig375
RMA4.75
FP50
sig475
RMA5.03
FP75
sig593
RMA5.30
FP100
sig795
RMA5.65
IFN Data       
Fraction Present (MAS5) 1230 20 30 49 84 165
Average Signal (MAS5) 1230 8 26 43 57 112
Fraction Present (RMA) 1641 53 92 143 218 334
Average RMA (RMA) 1641 92 161 205 252 346
  Unfiltered (total) FP>0
sig200
RMA5.05
FP25
sig265
RMA5.47
FP50
sig323
RMA5.70
FP75
sig400
RMA5.95
FP100
sig475
RMA6.17
Vitamin A Data       
Fraction Present (MAS5) 168 8 10 10 14 17
Average Signal (MAS5) 168 5 9 13 19 24
Fraction Present (RMA) 177 5 7 10 14 18
Average Signal (RMA) 177 13 17 22 28 31
  1. aFor comparison of different filters, thresholds were selected for each filter that removed similar number of probe sets. Types of filter: Fraction Present, 'FP' followed by threshold value; Average MAS5 signal, 'sig' followed by signal threshold; average RMA value, 'RMA' followed by threshold value. Total number of probe sets with p ≤ 0.001, followed by the number removed at each threshold (cumulative).
\