Transcription factors are key regulatory elements that control gene expression. The TRANSFAC database represents the largest repository for experimentally derived transcription factor binding sites (TFBS). Understanding TFBS, which are typically conserved during evolution, helps us identify genomic regions related to human health and disease, and regions that might be predictive of patient outcomes. Here we present a statistical analysis of all TFBS in the TRANSFAC database. Our analysis suggests that current definition of TFBS core regions in TRANSFAC should be re-examined so as to capture a more precise notion of "cores." We offer insight into more appropriate definitions of TFBS consensus sequences and core regions. These revised definitions provide a better understanding of the nature of transcription factor-DNA binding and assist with developing algorithms for de novo TFBS discovery as well as finding novel variants of known TFBS.

A combination of algorithms to search RNA sequence for the potential for secondary structure formation, and search large numbers of sequences for structural similarity, were used to search the 5'UTRs of annotated genes in the Escherichia coli genome for regulatory RNA structures. Using this approach, similar RNA structures that regulate genes in the thiamin metabolic pathway were identified. In addition, several putative regulatory structures were discovered upstream of genes involved in other metabolic pathways including glycerol metabolism and ethanol fermentation. The results demonstrate that this computational approach is a powerful tool for discovery of important RNA structures within prokaryotic organisms.

This paper focuses on the optimization of population and selection parameters of an evolutionary algorithm for similar RNA structure discovery. The effects of population settings such as the number of parents and number of offspring per parent, and method of selection (tournament vs. elitist) on the rate of convergence were investigated relative to a problem with a known RNA structure solution. The results indicate that proper setting of the number of parents and offspring can be used to increase the efficiency of the evolutionary process.

Transcription factors are key regulatory elements that control gene expression. Recognition of transcription factor binding site (TFBS) motifs in the upstream region of coexpressed genes is therefore critical towards a true understanding of the regulations of gene expression. The task of discovering eukaryotic TFBSs remains a challenging problem. Here, we demonstrate that evolutionary computation can be used to search for TFBSs in upstream regions of genes known to be coexpressed. Evolutionary computation was used to search for TFBSs of genes regulated by octamer-binding factor and nuclear factor kappa B. The discovered binding sites included experimentally determined known binding motifs as well as lists of putative, previously unknown TFBSs. We believe that this method to search nucleotide sequence information efficiently for similar motifs will be useful for discovering TFBSs that affect gene regulation.

Artificial neural networks (ANNs) can be utilized to generate predictive models of quantitative structure–activity relationships between a set of molecular descriptors and activity. Evolutionary computation provides a means to appropriately search for the set of weights and bias terms associated with artificial neural networks that minimize selected functions of the error between the actual and desired outputs. This method is demonstrated by evolutionary training of artificial neural networks capable of predicting anti-HIV activity for a set of 1-[(2-hydroxyethoxy)methyl]-6-(phenylthio)thymine (HEPT) derivatives. The results of this work further confirm the growing indication that evolutionary computation can outperform backpropagation as a method of artificial neural network training. The results also indicate the degree to which bias in the initial training and testing data can affect performance and the importance of bootstrapping.

RNA molecules fold into characteristic secondary and tertiary structures that account for their diverse functional activities. Many of these RNA structures, or certain structural motifs within them, are thought to recur in multiple genes within a single organism or across the same gene in several organisms and provide a common regulatory mechanism. Search algorithms, such as RNAMotif, can be used to mine nucleotide sequence databases for these repeating motifs. RNAMotif allows users to capture essential features of known structures in detailed descriptors and can be used to identify, with high specificity, other similar motifs within the nucleotide database. However, when the descriptor constraints are relaxed to provide more flexibility, or when there is very little a priori information about hypothesized RNA structures, the number of motif 'hits' may become very large. Exhaustive methods to search for similar RNA structures over these large search spaces are likely to be computationally intractable. Here we describe a powerful new algorithm based on evolutionary computation to solve this problem. A series of experiments using ferritin IRE and SRP RNA stem-loop motifs were used to verify the method. We demonstrate that even when searching extremely large search spaces, of the order of 10(23) potential solutions, we could find the correct solution in a fraction of the time it would have taken for exhaustive comparisons.