piSVM



1. INTRODUCTION


This program is developed for making prediction of piRNA within small RNA data. This method is appropriate for those species which have the transposon sequences. All the candidate small RNA sequences are matched to the transposon sequences, then we trained a SVM classifier to predict the piRNA, during the processing, 32-vectors triplets feature elements are used for training the SVM classifier.

2. Usage


2.1 Download the perl scripts and the transposon sequences.

2.2 Process the small RNA sequences by using the perl scripts step by step.

2.3 Make the prediction with the SVM classifier in MATLAB.

PS: this method needs the third-party software, they can be downloaded as follows:

  • Seqmap (http://www-personal.umich.edu/~jianghui/seqmap/)

  • RNAplex (http://www.bioinf.uni-leipzig.de/~htafer/RNAplex/RNAplex.html)

3. DOWNLOAD


All of the perl scripts and SVM models are free for all users but without any warranty.
Please contact Kai Wang at wangkai19890329@outlook.com for the download.