PSRna: Prediction of small RNA secondary structures based on reverse complementary folding method.

Area of Science:

• Computational biology
• Bioinformatics
• Molecular biology

Background:

• RNA secondary structures are vital for RNA molecule function.
• Predicting small RNA secondary structures is challenging due to data scarcity.
• Existing algorithms are not optimized for small RNA prediction.

Purpose of the Study:

• To develop a novel algorithm, PSRna, for accurate small RNA secondary structure prediction.
• To address the limitations of current methods in handling small RNAs.
• To improve the understanding of small RNA functions through precise structural predictions.

Main Methods:

• PSRna utilizes reverse complementary folding and characteristic hairpin loops specific to small RNAs.
• It employs dynamic programming to estimate the maximum number of base pairs and constructs a path matrix.
• Backtracking algorithms extract secondary structures, which are then filtered by minimum free energy for accuracy.

Main Results:

• PSRna demonstrates superior performance compared to RNAfold and RNAstructure.
• The algorithm achieves higher sensitivity, specificity, and Matthews correlation coefficient (MCC).
• Experimental validation on real data confirms the algorithm's effectiveness.

Conclusions:

• PSRna offers a significant advancement in small RNA secondary structure prediction.
• The algorithm's accuracy and efficiency make it a valuable tool for computational biology and bioinformatics.
• This work facilitates deeper insights into small RNA functions and mechanisms.