Artificial intelligence applied to omics data in liver diseases: Enhancing clinical predictions

Area of Science:

• Bioinformatics and Computational Biology
• Artificial Intelligence in Medicine
• Genomics and Precision Medicine

Background:

• Biotechnology advancements generate vast multi-omics data, requiring sophisticated computational tools.
• Artificial intelligence (AI) and machine learning (ML) are crucial for analyzing complex biological datasets.
• Accurate diagnosis and prognosis of non-malignant liver diseases are critical for effective patient management.

Purpose of the Study:

• To review the literature on AI models utilizing omics data for diagnosing and predicting outcomes in non-malignant liver diseases.
• To evaluate the impact of integrating omics data with clinical information on AI model performance.
• To assess the potential of AI-driven omics analysis for personalized medicine in liver disease.

Main Methods:

• Systematic literature review of studies employing AI/ML models on omics datasets (genomics, metabolomics, etc.) and clinical data.
• Analysis of 20 different AI models applied to diagnose, risk-stratify, and predict survival in non-malignant liver diseases.
• Comparison of model performance metrics, such as Area Under the Curve (AUC), with and without omics data integration.

Main Results:

• Integration of omics data generally improved AI model performance compared to using clinical data alone.
• Specific examples show drastic AUC improvements, e.g., from 0.87 to 0.99 with metabolomic data for NAFLD fibrosis staging.
• AI models using multi-omics and clinical data outperformed those using only clinical parameters or serum biomarkers for predicting disease progression and classification.

Conclusions:

• Omics data integration significantly enhances the predictive power of AI models for non-malignant liver diseases.
• AI models trained on multi-omics and clinical data hold substantial promise for improving diagnostic accuracy and prognostic prediction.
• These findings support the potential application of AI-driven omics analysis in developing personalized medicine strategies for liver disease patients.