Computational frameworks integrating deep learning and statistical models in mining multimodal omics data.

Area of Science:

• Biomedical data science
• Computational biology
• Health informatics

Background:

• Multimodal omics data analysis is crucial for health research.
• Traditional statistical methods have limitations with high-dimensional data.
• Deep learning excels at prediction tasks with omics data.

Purpose of the Study:

• To review state-of-the-art integrative frameworks for omics data analysis.
• To combine strengths of statistical methods and deep learning.
• To improve prediction accuracy, interpretability, and explainability.

Main Methods:

• Review of current integrative frameworks.
• Discussion of limitations and future directions.
• Focus on survival analysis, dimension reduction, regression, classification, feature selection, and learning transfer.

Main Results:

• Integrative frameworks offer improved prediction accuracy.
• These methods aim to enhance interpretability and explainability.
• The review covers diverse applications within omics data analysis.

Conclusions:

• Integrative frameworks represent a significant advancement in omics data analysis.
• Future research should focus on addressing limitations and expanding applications.
• Combining statistical and deep learning approaches is key for future discoveries.