Apis mellifera pre-miRNA prediction using decision tree based classifier

In this paper, we have used decision tree based classification approach to predict pre-miRNA from Apis mellifera (Honey bee). A set of 108 pre-miRNA from Apis mellifera with equal number of known pre-miRNA and non pre-miRNA is used for training the classifier. These 108 instances have 14 attributes that are derived from secondary structure data. The secondary structures are generated from pre-miRNA sequences from Apis meillfera database using the package RNAfold. We have applied information gain method for dimension reduction and attribute relevance analysis. This method reduces the dimension of the data set from 14 to 4 relevant attributes. The data set of resultant attributes is used for training the model for pre-miRNA prediction. Further the model is applied to two test data sets of 153 and 16 instances respectively for verification of prediction accuracy of the model. The precision and recall results from test data sets are encouraging and may facilitate miRNA biogenesis.