Figure 1From: Statistical implications of pooling RNA samples for microarray experimentsRelationships among type I error rate, sample size, effect size, and power with or without pooling. Note: n is the number of biological replicates per treatment group; c is the number of gene chips per group; a is the type I error rate; EQ means that samples are pooled with equal contribution; NE means samples do not contribute equally when pooled together (weights assigned randomly to each chip: 0.7, 0.2, and 0.1).Back to article page