Computational methods to predict long noncoding RNA functions based on co-expression network.

Area of Science:

• Genomics
• Bioinformatics

Background:

• Long noncoding RNAs (lncRNAs) are abundant in the mammalian transcriptome, but their functions are challenging to determine at scale.
• Current methods for lncRNA functional characterization through lab experiments or structure prediction are limited.

Purpose of the Study:

• To develop and present a computational pipeline for the large-scale functional annotation of lncRNAs.
• To address the deficit in understanding lncRNA functions by utilizing gene co-expression networks.

Main Methods:

• Constructed a coding-noncoding gene co-expression network using gene expression profiles, specifically in prostate cancer.
• Nodes represent protein-coding genes or lncRNAs; edges signify co-expression based on correlation coefficients.
• Employed model-based and hub-based sub-networks for predicting lncRNA functions.

Main Results:

• Successfully developed a computational pipeline for lncRNA functional annotation.
• Demonstrated the pipeline's application in constructing a co-expression network for prostate cancer data.
• Showcased the prediction of lncRNA functions through network analysis.

Conclusions:

• The developed computational pipeline enables large-scale functional annotation of lncRNAs.
• Co-expression network analysis provides a viable approach for predicting lncRNA functions, particularly in disease contexts like prostate cancer.