The F (or ‘Felgner’) plot is basically a comparison of the antibody profiles of two populations. It is represented by a clustered column or bar chart of the mean signal intensities of each group, which is overlaid with p-values.
The downloadable Excel file below walks through the steps. For the bar chart, raw data subtracted of the median of the IVTT control spots looks best, although raw data divided by the median IVTT control (FOC) can also be used. Although log transformed data is required to calculate the p-values (see below), log-transformed data is generally not as visually impressive when represented as bar charts.
To calculate the p-values for overlaying onto the bar chart, use log-transformed FOC data. Set a floor of 100 for the raw data and then normalize by dividing the IVTT protein spot intensity by the sample specific median of the multiple IVTT control spots, then taking the base-2 logarithm of the ratio. A two-sided T-test is performed for each antigen between the two groups of interest, followed by adjustment of the p-value for false discovery using the Benjamini-Hochberg (BH) method. The attached example shows the work-around for BH correction that can be performed in Excel, although for publication a more acceptable stats package is recommended for both the p-value calculation and correction. (Method from Arlo Randall, Antigen Discovery Inc).
Huw's Proteome Microarrays User Manual 