Microarray technologies have proven useful in studies of cancer and their use is now widespread. To optimize information from such studies, good experimental design is critical. Unfortunately, experimental design has received relatively little attention and a number of important questions remain open. One question is whether or not to pool messenger RNA (mRNA) samples across subjects. There are good reasons both for and against pooling; and as a result both types of designs are used in practice. The type of design used will certainly affect data analysis and inference; but the exact effects are not yet completely known as pooling has only recently been studied in the context of microarray experiments. Preliminary statistical calculations have shown that pooling mRNA samples across subjects can reduce the effect of biological variability and in so doing reduce the number of arrays required in an experiment. The extent of the reduction depends on ratios of experimental variances. The overall goal of this research project is to obtain a unique data set from which we can experimentally validate assumptions made in statistical calculations, estimate variance ratios in an experimental population, and provide guidelines on how to easily extend the results to other populations.
| Kendziorski, C; Irizarry, R A; Chen, K-S et al. (2005) On the utility of pooling biological samples in microarray experiments. Proc Natl Acad Sci U S A 102:4252-7 |
