A Comprehensive Review of Deep Learning Applications with Multi-Omics Data in Cancer Research.

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Deep learning (DL) and multi-omics data integration are transforming biological system analysis, especially in cancer research.
• DL facilitates high-dimensional data analysis, uncovering novel disease mechanisms and biomarkers for enhanced patient care.

Purpose of the Study:

• To provide a comprehensive review of recent deep learning models applied to cancer genomics.
• To focus on key applications including cancer type classification, driver gene identification, survival analysis, and drug response prediction.

Main Methods:

• Systematic review of methods published since 2020.
• Analysis of deep learning model architectures, datasets, and innovations.
• Explanation of foundational machine learning and deep learning concepts and multi-omics data characteristics.

Main Results:

• Recent deep learning models demonstrate significant potential in cancer genomics applications.
• Key innovations in model architectures and data integration are highlighted.
• The review covers advancements in classification, driver gene identification, survival analysis, and drug response prediction.

Conclusions:

• Deep learning integration with multi-omics data offers powerful tools for advancing cancer research.
• These advancements are crucial for developing improved diagnostic and therapeutic strategies.
• The review serves an interdisciplinary audience interested in computational approaches to cancer biology.