GANN: Genetic algorithm neural networks for the detection of conserved combinations of features in DNA

GANN (available at http://bioinformatics.org.au/gann webcite) is a machine learning tool for the detection of conserved features in DNA. The software suite contains programs to extract different regions of genomic DNA from flat files and convert these sequences to indices that reflect sequence and structural composition or the presence of specific protein binding sites. The machine learning component allows the classification of different types of sequences based on subsamples of these indices, and can identify the best combinations of indices and machine learning architecture for sequence discrimination. Another key feature of GANN is the replicated splitting of data into training and test sets, and the implementation of negative controls. In validation experiments, GANN successfully merged important sequence and structural features to yield good predictive models for synthetic and real regulatory regions.GANN is a flexible tool that can search through large sets of sequence and structural feature combinations to identify those that best characterize a set of sequences.The minimal requirement for transcriptional activation is recruitment of an RNA polymerase complex to a promoter sequence of DNA upstream of an open reading frame (ORF). Most genes are also potentially under the control of DNA-binding regulatory proteins or transcription factors that can activate or silence transcription. In bacteria, activator and repressor proteins bind to operator sequences that are typically found near the promoter, and promoter specificity is typically conferred through the sigma subunit of RNA polymerase, which binds the promoter directly [1]. Eukaryotic transcription factors interact with DNA within the promoter, and are responsible for recruitment of the RNA polymerase complex [2]. Regulatory proteins also bind to conserved sites near the promoter region, as well as to enhancers that can be far (> 10 000 nucleotides) upstream or downstream of the promoter. In all domains of l