The program tests microarray data for significance. It calculates pvalues and adjusted pvalues for every gene.
The procedure works as follows:
First, the logarithms of the ratios are calculated. The logarithms of the ratios follow a normal distribution. Then, a standard ttest is performed on the transformed values. This yields a confidence interval for the true value of the ratio and a pvalue for every gene.
From the list of pvalues adjusted pvalues are calculated. The adjusted pvalues give estimations for the false discovery rate and the family wise error rate.
Nice things you also get from this:
 A histogram and
 a QQPlot of the Log ratios.
Literature
Adjusted pvalues:

The false discovery rate: Storey JD. (2002) A direct approach to false discovery rates.
Journal of the Royal Statistical Society, Series B, 64: 479498

The more conservative false discovery rate, reported in the BH column of the results of the analysis:
Y. Benjamini and Y. Hochberg (1995).
Controlling the false discovery rate: a practical and powerful approach to multiple testing.
J. R. Statist. Soc. B. Vol. 57: 289300.

The adjusted pvalue (control of FDR) in the BY column:
Y. Benjamini and D. Yekutieli (2001).
The control of the false discovery rate in multiple hypothesis testing under dependency.
Annals of Statistics.

The adjusted pvalue (control of FWER) in the Hochberg column:
Y. Hochberg (1988).
A sharper Bonferroni procedure for multiple tests of significance,
Biometrika. Vol. 75: 800802.

The adjusted pvalue (control of FWER) in the Holm column:
S. Holm (1979).
A simple sequentially rejective multiple test procedure.
Scand. J. Statist.. Vol. 6: 6570.