TY - CONF T1 - Characterizing RNA Secondary-Structure Features and Their Effects on Splice-Site Prediction T2 - Data Mining Workshops, 2007. ICDM Workshops 2007. Seventh IEEE International Conference on Y1 - 2007 A1 - Dogan,R. I A1 - Getoor, Lise A1 - Wilbur,W. J KW - algorithm;gene KW - biophysics;molecular KW - composition;splice-site KW - compounds;sequences; KW - computing;genetics;macromolecules;molecular KW - configurations;organic KW - features;feature KW - generation KW - model;splice-site KW - molecules;RNA KW - neighborhood;three-dimensional KW - pairing KW - RNA KW - secondary-structure KW - sequence KW - shape;biology KW - splicing;nucleotide KW - tendency;pre-mRNA AB - RNA molecules are distinguished by their sequence composition and by their three-dimensional shape, called the secondary structure. The secondary structure of a pre-mRNA sequence may have a strong influence on gene splicing. In our previous work, we showed that a splice-site model employing sequence features built using our feature generation algorithm was very effective in predicting splice sites. The generated sequence features also contained biologically relevant features. In this paper, we extend the feature generation algorithm to construct secondary-structure features. These features capture the nucleotide pairing tendency in the splice-site neighborhood. We extend the splice-site model to include both pre-mRNA sequence and structure characteristics. The new model significantly outperforms the sequence-based features model. The identified secondary-structure features capture biologically relevant signals such as splicing silencers. We also found these signals to prefer specific regions around the splice-site neighborhood and we detail their preference. JA - Data Mining Workshops, 2007. ICDM Workshops 2007. Seventh IEEE International Conference on M3 - 10.1109/ICDMW.2007.119 ER -