NPI-RGCNAE: Fast Predicting ncRNA-Protein Interactions Using the Relational Graph Convolutional Network Auto-Encoder.

Area of Science:

• Bioinformatics
• Computational Biology
• Molecular Biology

Background:

• Non-coding RNAs (ncRNAs) are crucial regulators of biological processes.
• Understanding ncRNA-protein interactions is key to deciphering ncRNA functions.
• Experimental identification of these interactions is resource-intensive.

Purpose of the Study:

• To develop an efficient computational method for predicting ncRNA-protein interactions.
• To address the limitations of experimental approaches in identifying these crucial molecular partnerships.

Main Methods:

• Developed NPI-RGCNAE, a novel method utilizing a Relational Graph Convolutional Network Auto-Encoder.
• Employed a reliable negative sample selection strategy.
• Applied a Relational Graph Convolutional Network encoder and DistMult decoder for prediction.

Main Results:

• NPI-RGCNAE demonstrated comparable performance to state-of-the-art methods via 5-fold cross-validation.
• The method achieved high accuracy in predicting ncRNA-protein interactions.
• Achieved prediction with significantly reduced computational time, requiring less than 10% of the training time of existing methods.

Conclusions:

• NPI-RGCNAE provides an accurate and highly efficient computational approach for predicting ncRNA-protein interactions.
• The method is particularly suitable for large-scale datasets due to its speed.
• Offers a valuable tool for advancing research in ncRNA function and regulation.