Computational Identification of piRNAs Using Features Based on RNA Sequence, Structure, Thermodynamic and...

Area of Science:

• Bioinformatics
• Genomics
• Molecular Biology

Background:

• PIWI-interacting RNAs (piRNAs) are small non-coding RNAs (ncRNAs) crucial for gene regulation and genome stability.
• Despite their importance, many piRNAs remain unidentified due to limitations in current predictive tools.

Purpose of the Study:

• To develop an advanced computational framework, piRNAPred, for accurate piRNA identification.
• To enhance the discovery of novel piRNAs and elucidate their biological roles.

Main Methods:

• Developed piRNAPred, an integrated framework using k-mer composition, secondary structure, and physicochemical properties.
• Trained machine learning models, notably Support Vector Machine (SVM), on a curated dataset of verified piRNAs and non-piRNA sequences.

Main Results:

• Achieved high prediction accuracy (98.60%) with a Mathews correlation coefficient (MCC) of 0.97 and ROC of 0.99 during 10-fold cross-validation.
• Successfully reduced feature space dimensionality using an attribute selection classifier.

Conclusions:

• piRNAPred demonstrates superior performance compared to existing piRNA prediction tools.
• This framework will facilitate the expansion of the known piRNA repertoire and deepen our understanding of piRNA functions.