An SVM-based approach for discovering splicing junctions with RNA-Seq

Because of the complicity of RNA splicing, discovering splicing junctions and aligning RNA-Seq reads with high sensitivity and specificity is still a challenging problem. In this paper, we propose an SVM-based approach for discovering splicing junctions with RNA-Seq data. The key idea of our approach is combining multiple sources of information contained in sequence data in discovering exon-intron boundaries. Results on simulated data suggest that our approach works well.