Software for Analysis of Gene Regulatory Sequences by Knowledge Discovery Methods

Abstract

Application of knowledge discovery techniques to search for and analysis of regularities in the context signals in DNA sequences involved in transcription regulation is described. The software developed allows for interactive construction and visualization of complex signals for samples of DNA regulatory sequences and evaluation of their statistical significance. The complex signals in DNA comprise simple context signals (oligonucleotides), the signals with specified localizations in promoters, predicted and experimentally detected transcription factor binding sites, and other user-specified signals. The regularities are statistically assessed by first-order logic with probabilistic estimates. The computer tool developed was tested using samples of nucleotide sequences of eukaryotic gene promoters extracted from databases according to tissue specificity, type of gene regulation, or joint expression in a functional system. A set of conservative context signals relating the nucleotide sequences and gene functional class was found. The software is available by request to the corresponding author.