MA Plot

An MA plot visualizes the relationships between the log ratio and mean values of two variables:

•

The “M” refers to minus in the log scale. The log ratios of the two measurements are plotted on the vertical (y) axis.

•

The “A” refers to average in the log scale. The mean values of the two measurements are plotted on the horizontal (x) axis. The A component serves as a common data range reference (for example, the average across all arrays). Many processes in JMP Genomics use the average as the common reference A component, making the MA plots visually comparable through shared x-coordinates, unless you override this default.

In the context of microarray expression data, MA plots are used to examine the intensity measurements resulting from two channels. These channels can be drawn from either:

•

One chip from a two-color platform containing red (R) and green (G) channels, or

•

Two or more chips from a one-color platform. (We can call the intensity measurements from chip 1 R, and those from chip 2 G.)

Regardless of the data source, M and A are defined as:

•

M = log2(R / G), and

•

A = (1/2) * log2(R * G)

A standard assumption in some microarray experiments is that few genes should exhibit a change in expression, meaning that most points on M should be close to zero (0). Under this assumption, wild deviations from zero suggest that the data should be further normalized before additional statistical analyses are conducted.

Example MA plots (in contour plot form) are shown below.

These plots reveal the effect of KDMM normalization on sample data intensity ratio by average intensity.

Tip: You can obtain direct (non-contour) MA plots by unchecking the Apply normalized data for MA plot box in the KDMM Normalization and TMM Normalization process dialogs.