In principle, we can state that the causal relationship between a regulator and target gene exists if we see the effect of gene deletion; however, we cannot say the opposite: that if the effect is not observed, then the connection does not exist. A relationship may exist but may not be observable. Building conclusions about network topology by drawing links between genes whose causal relationship was discovered by mutation experiments can lead to incomplete and sometimes even incorrect connections. Without including the dynamics of the system into the network, its functional properties cannot be studied and interpreted correctly. Logical interpretation of observations can be completely wrong when the regulation is complicated and includes a cascade of reactions. Therefore,Cefetamet pivoxil HCl to discover regulatory relationships in genetic networks, one cannot rely solely upon static data but must also consider the dynamics of the network. Using a mathematical description of regulation of gene expression and a procedure for computation of its parameters from experimental data, it was possible to construct a complete numerical model of the genetic network of yeast cyclins active during cell cycle. The model was able to fully describe the kinetics of gene expression of any gene of the reconstructed network coherently with the measured gene expression profiles. The model allowed the simulation of a situation when genes in the topmost level of the regulatory cascade were deleted, which simulated experimental gene deletion. Influence of such deletion on the change in the expression profiles of other genes of the network was analyzed. The virtual gene deletion showed that in more complicated cascades of regulation, with many genes in between the deleted gene and the target gene, the result of gene deletion is quite unpredictable and,Butylhydroxyanisole in several cases, the absence of the deleted gene can be compensated within the cascade. This compensation means that even if there is a causal relationship between the deleted gene and the target gene, it may not be discovered by the mutation experiment. Conclusions drawn from the dynamic model of the cyclins genetic network can be criticized because they are not experimentally verified. Although this model has been verified by comparison with previously measured experimental data, this point cannot be ignored. If the network model is wrong in particular connections, then parameters that were computed to fit experimentally measured expression profiles would be wrong as well; thus, the results of virtual gene deletion could also be wrong or, at least, altered. A crucial point of this paper is that the response to mutation of genes in the topmost layer has highly unpredictable impact on the genes lower in the causal cascade and that the effect of mutation can disappear in the cascade of reactions. This statement remains unchanged even if the model is, in certain cases, wrong. An error would influence interpretation of a particular network of cyclins, what is indeed, with currently available data possible, but would not change the basic conclusion that the network dynamic is essential in the interpretation of its biological function. Another point of discussion is the linearity of the relationship between mRNA and final protein concentration. Recent studies show that, as measured by microarrays or qPCR, almost 50% of genes exhibit a linear relationship whereas others are either posttranslationally modified to alter their activity or their relationship is nonlinear for other reasons. For this reasons we excluded from our analysis genes that are known to be controlled postranscriptionally.