Skip to main content
Figure 1 | BMC Bioinformatics

Figure 1

From: A comparative review of estimates of the proportion unchanged genes and the false discovery rate

Figure 1

Boxplots of the results from eight methods for estimating p0 using simulated independent data. A simulation model of real-life microarray data was used to give data where the expected proportion of unchanged genes was set at 50, 60, 70, 80, 90, 95 or 99%. The central box of the boxplot represents the data between the 25% and 75% percentiles with the median represented by a line. Whiskers go out to the extremes of the data, and very extreme data appear as points. The abscissa shows the expected proportion unchanged and ordinata the estimate. Legends : A – BUM;B – SPLOSH; C – smoother; D – bootstrap; E – SEP; F – LSL; G – mgf; H – PRE. A and B underestimate in the low range and in the high range, respectively; C and D appear very similar with a sound overestimation and some high variance; E appears stable and reliable; F gives a stable and quite conservative estimate; G overestimates and varies very little; G is stable and gives a small degree of overestimation.

Back to article page