A recent breakthrough in application of experimental methods to biomedicine not only has already resulted in accumulation of massive amounts of data, but also created new challenges in the field of data mining. This paper addresses the issue by combining advanced data mining techniques with the novel application of systems biology t o study axonal regeneration and neurogenesis. To obtain the data for analysis, a series of biological experiments are conducted, in which the rat pheochromocytoma cell PC-12 was used as an appropriate model for neuronal differentiation. The resulting DNA microarray data set is studied using a combination of methods—a statistical procedure for feature selection together with a dimensionality reduction technique. First, we apply feature selection, which can be based either on the Wilcoxon rank-sum test, or on the two-sample t-test, depending on the statistical properties of the data. Next, we utilized an efficient dimensionality reduction procedure called correspondence analysis to obtain a two-dimensional projection of the data, which allows us to perform a visual exploration of the patterns as well as to select the features corresponding to the top down-regulated and up-regulated genes.