Multi-omics strategies for biomarker discovery and application in personalized oncology
Ziming Jiang, Haoxuan Zhang, Yibo Gao, Yingli Sun · 2025 · 98 citationsRead the paper
Multi-omics strategies, integrating genomics, transcriptomics, proteomics, and metabolomics, have revolutionized biomarker discovery and enabled novel applications in personalized oncology. Despite rapid technological developments, a comprehensive synthesis addressing integration strategies, analytical workflows, and translational applications has been lacking. This review presents a comprehensive framework of multi-omics integration, encompassing workflows, analytical techniques, and computational tools for both horizontal and vertical integration strategies, with particular emphasis on machine learning and deep learning approaches for data interpretation. Recent applications of multi-omics have yielded promising biomarker panels at the single-molecule, multi-molecule, and cross-omics levels, supporting cancer diagnosis, prognosis, and therapeutic decision-making. However, major challenges persist, particularly in data heterogeneity, reproducibility, and the clinical validation of biomarkers across diverse patient populations. This review also highlights cutting-edge advances in single-cell multi-omics and spatial multi-omics technologies, which are expanding the scope of biomarker discovery and deepening our understanding of tumor heterogeneity. Finally, we discuss the integral role of multi-omics in personalized oncology, with a particular focus on predicting drug responses and optimizing individualized treatment strategies, supported by real-world clinical practice cases. By bridging technological innovations with translational applications, this review aims to provide a valuable resource for researchers and clinicians, offering insights into both current methodologies and future directions for implementing multi-omics data in biomarker discovery and personalized cancer care.
What the full text shows
Deployed in practice- Multi-omics integration of genomics and transcriptomics identifies key markers for EGFR mutations and ALK gene fusions in NSCLC, enabling targeted therapy with erlotinib and crizotinib. (Non-small cell lung cancer (NSCLC))
- Integration of genomics, proteomics, and transcriptomics improves identification of HER2-positive breast cancer patients for treatment with trastuzumab and pertuzumab. (HER2-positive breast cancer)
- Multi-omics analysis of tumor mutational burden (TMB) and neoantigens predicts response to PD-1 and CTLA-4 inhibitors in melanoma. (Melanoma)
- A 10-metabolite plasma signature provides superior diagnostic accuracy compared to conventional tumor markers. (Gastric cancer)
- Large-scale sequencing (MSK-IMPACT) reveals that a significant portion of tumors harbor actionable alterations. — 37% of tumors (General tumor samples)
- TMB is an FDA-approved predictive biomarker for pembrolizumab treatment. (Solid tumors)
- Setup
- Review of multiple datasets including TCGA, PCAWG, MSK-IMPACT, CPTAC, and specific databases like DriverDBv4 (24,000 patients) and GliomaDB (4,303 patients).
- How
- The paper reviews a framework for horizontal (intra-omics) and vertical (inter-omics) integration using machine learning, deep learning, and batch correction algorithms (e.g., Seurat, Harmony). It synthesizes clinical case studies and evaluates existing multi-omics databases and quality control protocols.
- Limitations
- Risk of false positives due to statistical overfitting and technical biases in high-dimensional data. · Requirement for extensive functional validation of large numbers of candidate biomarkers. · Lack of unified standards for storing and managing multi-omics databases. · Difficulty in tracing some omics measurements to International System of Units (SI units).
Summarized by Gemma 4 from the open-access full text · 14 Oct 2026. Read it