2lpiRNApred

2lpiRNApred is a computational tool to predict and analyze Piwi-interacting RNAs (piRNAs), crucial for transposon silencing, germ cell development, and maintaining genome integrity. Recognizing the essential roles of piRNAs in various biological processes and their emerging significance in tumorigenesis, 2lpiRNApred employs a novel two-layered integrated classifier algorithm. This algorithm excels in identifying piRNAs and determining their potential function in inducing target mRNA deadenylation—a critical post-transcriptional modification influencing gene expression and potentially involved in cancer development.

To enhance the tool's predictive accuracy and handle the complexity of piRNA functions, 2lpiRNApred incorporates a new feature selection algorithm based on Luca fuzzy entropy and the Gaussian membership function (LFE-GM), aimed at reducing feature dimensionality. The tool integrates five feature extraction strategies—Kmer, General parallel correlation pseudo-dinucleotide composition, General series correlation pseudo-dinucleotide composition, Normalized Moreau-Broto autocorrelation, and Geary autocorrelation—coupled with two types of classifiers: Sparse Representation Classifier (SRC) and a support vector machine with Mahalanobis distance-based radial basis function (SVMMDRBF).