Clinical Multi-Omics: What It Means in Cancer Care

Clinical multi-omics — the integration of genomics, transcriptomics, proteomics, metabolomics and epigenomics — is moving from research into everyday cancer care. By combining several molecular layers, clinicians can build a fuller picture of a tumour than any single test can provide.

What is clinical multi-omics?

Clinical multi-omics is the practice of analysing multiple types of molecular data from a patient — DNA, RNA, proteins, metabolites and epigenetic marks — to understand how a disease behaves. In oncology, this means looking at a tumour from several angles at once, which can help guide more personalised treatment decisions.

Better disease prediction with clinical multi-omics

Recent studies have shown that, compared with clinical data alone, the integration of genomics, transcriptomics, proteomics, metabolomics, and epigenomics provides better prediction for the onset of diseases.

Matching patients to treatment

Multiomics improves the ability to recognize patients that are most likely to benefit from a given treatment, especially oncology and immunotherapy.

AI and data integration in multi-omics

Artificial intelligence, graph-based learning and network-based models are becoming the key methods for multiomics data integration and for the identification of novel biomarkers, disease subtypes, and therapeutic targets.

Cancer research at single-cell resolution

Cancer research is aided by the ability to study individual cells in their native (!) tissue locations.

Making multi-omics more accessible

Multiomics is transforming the ability to study individual cells in their native tissue locations, which provides greater insights into the heterogeneity of tumors, the progression of disease, and treatment resistance. Many research facilities are adopting automated workflows, making this research easier to access.

Some clinical results

• Predicting which patients are likely to develop invasive lung cancer (and potentially avoiding unnecessary biopsy procedures).
• Predicting which patients are likely to develop diabetes and cardiovascular diseases.
• Predicting which patients are likely to benefit from immunotherapy for liver cancer.
• Predicting which patients with rare disorders are most likely to benefit from novel integrated genomic and transcriptomic sequencing.

What's next?

Hurdles still exist for multiomics to occur routinely in clinics, including:

• Integration of data at different molecular levels
• Incomplete datasets
• Problems with reproducibility and consistency
• Validation in large clinical cohorts
• Costs and practicality

Although clear problems and barriers exist, we are highly optimistic. We are confident the next best thing in history won't be just gathering and analyzing data like never before — it will be AI. Precision medicine will move beyond relying on an individual's DNA and will depend upon multiple biological layers for accurate diagnosis, more effective therapeutic options, and tailored care.

Multiomics is in its final frontier of testing. Systems and services will focus on improving this cornerstone of research faster and more efficiently than anything we have ever seen.

Further reading: Going beyond genomics in precision oncology