CisPi: a transcriptomic score for disclosing cis-acting disease-associated lincRNAs

Area of Science:

• Genomics
• Computational Biology
• Cancer Research

Background:

• Long intergenic noncoding RNAs (lincRNAs) are emerging as crucial biomarkers in cancer biology.
• Predicting cis-acting transcription of lincRNAs from enhancers is challenging due to the lack of specific hallmarks.
• Understanding lincRNA-gene interactions is vital for deciphering regulatory mechanisms in diseases.

Purpose of the Study:

• To introduce a novel transcriptomic method, CisPi, for quantifying the association between lincRNAs and local gene expression.
• To develop a computational model for predicting risk-dependent lincRNAs and identifying active enhancers.
• To discover novel clinical biomarkers and prognostic indicators in neuroblastoma.

Main Methods:

• Developed CisPi, a metric quantifying mutual information between lincRNAs and local gene expression in response to perturbation.
• Utilized a side-by-side analytical pipeline to analyze lincRNAs, especially those with low read counts in RNA-Seq data.
• Applied the method to neuroblastoma to identify risk-dependent lincRNAs and their regulatory roles.

Main Results:

• CisPi successfully identified risk-dependent lincRNAs, revealing active enhancers and neuroblastoma susceptibility loci.
• Prioritized lincRNAs demonstrated significant prognostic value in neuroblastoma.
• Predicted target genes of these lincRNAs also inherited prognostic significance, highlighting their clinical relevance.

Conclusions:

• RNA-Seq data, analyzed with our methodologies, is sufficient for identifying disease-associated lincRNAs.
• CisPi offers a powerful tool for discovering lincRNA biomarkers and understanding gene regulation, even without explicit enhancer hallmarks.
• The approach has broad applicability in various biological contexts and diseases.