miRNA prediction using computational approach

The total number of miRNAs in any species even completely sequenced is still an open problem. However, researchers have been using limited techniques for miRNA prediction. Some of the prediction models are developed using different sets of features of pre-miRNA from model organisms. We also focused on exploring dominating set of features using principal component analysis and Info gain in our earlier work, with the assumption that it will reduce complexity of the model and may increase prediction accuracy. In this chapter, we have used attribute relevance analysis techniques for selecting essential attributes based on their relevance. We have explored dominating feature extraction using different machine learning techniques from a set of known mature miRNA sequences. The results are encouraging since the essential attributes selected here are biologically significant. These attributes can be used in deriving rules for miRNA identification.