Elucidation of dynamic microRNA regulations in cancer progression using integrative machine learning

Area of Science:

• Genomics
• Computational Biology
• Machine Learning

Background:

• Advanced genomics tools generate vast data on gene regulation (transcription factors, microRNAs, epigenetics).
• Computational models represent gene regulatory mechanisms but face challenges with complex, dynamic biological systems like cancer.
• Heterogeneous data growth in cancer research necessitates scalable and integrative modeling approaches.

Purpose of the Study:

• To develop an integrative machine learning approach for inferring multifaceted gene regulations in cancer, with a focus on microRNA regulation.
• To identify conditional microRNA-mRNA interactions across different cancer stages using a supervised deep learning model.
• To analyze microRNA-mediated dysregulation in human breast cancer progression.

Main Methods:

• Integrative machine learning approach combining data integration and graphical model fusion.
• Supervised deep learning model for identifying conditional microRNA-mRNA interactions.
• Case study on human breast cancer across four progressive stages.

Main Results:

• Distinct gene regulatory networks identified for four progressive stages of human breast cancer.
• Functional analysis revealed significant changes in cancer hallmarks and novel signaling/metabolic processes due to microRNA dysregulation.
• Shed light on the regulatory roles of microRNAs in breast cancer progression.

Conclusions:

• The integrative model serves as a robust tool for understanding key regulatory characteristics in complex biological systems.
• Highlights the critical role of microRNAs in the progression of breast cancer.
• Suggests potential for novel therapeutic targets by understanding stage-specific regulatory networks.