is a machine learning algorithm based on SVM (support vector machine) that serves as a post-processing filter for the miRNA:target duplexes predicted by softwares such as miRanda
. For an input duplex, MiRTif outputs an SVM score indicating whether the input is likely to be a positive interaction.The SVM is trained with validated positive and negative miRNA:target interactions, obtained from TarBase
. These interactions possess the well-accepted miRNA:target binding properties such as strong seed complementarity and high binding energy, and are believed to be authentic. Therefore, MiRTif is only applicable to interactions obtained from other prediction software.
Users have two options to predict a miRNA:target interaction. You can input duplexes one at a time or upload a text file containing the list of the duplexes. It takes a few seconds to process a duplex in general. Nevertheless querying a long list of duplexes may require some time, the results can be emailed to you when finished.
manuscript in preparation
Selected informative features
MicroRNAs in the training set