riboswitch prediction tool
Riboswitches are a fragment of mRNA found along the 5’ untranslated region (UTR) found mostly in the bacterial genome. They act as regulatory elements by functioning like switches that turn the expression of genes on or off in response to binding of a metabolite to them. Thus, the binding of the small molecule to the riboswitch changes both the tertiary structure and the functionality of the RNA. There are several computational tools developed in the past that predict the riboswitch like element in the sequence submitted. Each tool has its own limitations such as query size, motif quality and so on which resulted in some of the potential riboswitch like elements to go undetected. In the present study, a motif-based pattern searching model has been proposed for the prediction of riboswitches and classifying them. This method uses mono, bi, tri and quad nucleotide sequence motif-based approach for training SVM to improve the accuracy of the output. The tool has been developed in Python based on sequential patterns. The proposed tool is able to predict 37 types of riboswitches with sensitivity ranging from 0.91 to 1.00 and the specificity falling in the range of 0.991 to 1.00.