Sample Pooling for Microarray Analysis: A Statistical Assessment of Risks and Biases Pooling of RNA samples isolated from

tissue is a strategy that can be imple-mented in microarray experiments when the amount of sample RNA is limiting, or when variation across samples must be reduced. However, it is widely documented that important gene expression information is lost during pooling. This Technical Note qualifies and quantifies the effects of pooling on biological conclusions derived from gene expression data. •Many transcripts identified as significantly changed in individual samples were not identified in the pooled samples. All information of small, but statistically significant, changes in expression was lost in the pooled samples. • It is recommended that researchers use non-pooled (individual) samples in order to identify statistically sig-nificant changes in gene expression