Global analysis of N6-methyladenosine functions and its disease association using deep learning and network-based...

Area of Science:

• Epigenetics and RNA Biology
• Computational Biology and Bioinformatics
• Cancer Genomics

Background:

• N6-methyladenosine (m6A) is the most prevalent mRNA modification, implicated in mRNA metabolism and various diseases, including cancer.
• The precise control mechanisms and disease relevance of m6A regulation remain largely unelucidated.
• Understanding m6A's role is crucial for advancing cancer research and therapeutic strategies.

Purpose of the Study:

• To develop and validate a computational framework for predicting m6A-regulated genes and their associated diseases.
• To identify key genes and pathways influenced by m6A modification in human samples.
• To uncover novel m6A-associated diseases and their underlying molecular mechanisms.

Main Methods:

• Development of Deep-m6A, a deep learning model for single-base resolution detection of condition-specific m6A sites from MeRIP-Seq data.
• Implementation of Hot-m6A, a network-based pipeline utilizing Protein-Protein Interaction (PPI) and gene-disease networks to prioritize functional m6A genes and diseases.
• Application of the computational scheme to 75 human MeRIP-seq samples.

Main Results:

• Identification of 709 functionally significant m6A-regulated genes and nine enriched subnetworks.
• Functional enrichment analysis revealed m6A targets critical genes in transcription, cell organization, proliferation, and cancer pathways (e.g., Wnt pathway).
• Prioritization of five significantly associated diseases, including leukemia and renal cell carcinoma.

Conclusions:

• The proposed computational scheme effectively identifies m6A-regulated genes and associated diseases.
• m6A plays a significant role in fundamental biological processes and cancer development.
• These findings provide valuable insights into m6A regulatory functions and disease implications, opening avenues for future research.