[Problems in using methods of analyzing gene expression in studying the aging process]

Aging processes are now actively investigated on the base of analysis of changes in gene expression. Many of the standard methods (such as hierarchic clustering, method of K-means, SOM (Self-organizing maps), Gene Shaving, etc.) do not take into account the specifics of time sets. In the article the new approach of the analysis and clustering of genes based on the processing of time increments in their expressions is presented. Genes in the resulting groups behave concertedly with any chosen "pilot" gene or simply together. The methods of clustering were tested on the data of gene expressions of C. elegans in aging. In the article the specific claims of believability of the experiments are discussed.