Using a Support Vector Machine to identify pre-miRNAs in soybean (Glycine max) introns

MicroRNAs (miRNAs) are small Ribonucleic Acid (RNA) molecules ~18-22 nucleotides (nt) in length that regulates gene expression in animals, plants and viruses. Due to its small size and occurrence in different development stages of organisms, the experimental identification of miRNAs becomes difficult, and computational approaches are being developed in order to precede and guide biological experiments. This paper describes our approach based on a Support Vector Machine (SVM) algorithm to identify miRNA's precursor (pre-miRNA) in soybean (Glycine max) transcript introns, that was developed using a secondary structure predictor of pre-miRNAs sequences to establish the feature set for training, testing and validation phases of SVM algorithm.