Modeling mRNA splicing by exon definition in silico

点击翻译按钮获取中文摘要

Interpretation of the effects of variation in genes is among the most challenging and important problems to be addressed in deciphering the sequences of complete genomes. My laboratory has developed bioinformatic theory, software, and applications to understand and predict the consequences of DNA sequence changes. This framework uses information theory to relate differences between the strengths of interactions between proteins and DNA or RNA to their effects on gene expression. Information theory can be used to detect and quantify signals in DNA or RNA that are recognized and bound by proteins in the cell. This project will develop and evaluate mathematical models for the natural cellular process to identify elemental protein-coding units of genes, termed exons, from the sequences of unprocessed RNA transcripts. Correct processing of these transcripts is called mRNA splicing. We will use information theory to define these signals, and then combine multiple signals within individual exons to distinguish and quantify correctly processed mRNA transcripts from potential decoys. The predictions will be compared with published sequence databases of expressed genes to determine the accuracy of these models. The models, methods, and software developed in this will be useful for predicting gene expression patterns in humans and other species. These resources will eventually be used in studies of common sequence variants that are predicted to affect normal mRNA splicing.