Reliable normalization

The use of at least two internal controls is recommended, as described above. If these correlate well, then additional ICs may be unnecessary (Figure 6.4A). However, if a different profile is found, then it is recommended that additional controls are used. Given that normalization can have dramatic effects on the results of relative quantification of gene expression, accurate normalization ensures that researchers do not report artifacts of normalization which may be misleading. Low expression of an internal control in just one sample can result in the normalized expression appearing erroneously higher. Such outliers can have dramatic effects when small sample sizes are used, and are capable of producing quite misleading data.

The use of additional metrics, such as gene stability index (Vandesompele et al., 2002) or the variance of IC expression within samples and across IC genes as described above (Figures 6.4B and 6.4C) enable a quantitative approach to IC selection.

0 0

Post a comment