Genomics has provided actionable clinical information, helping diagnose rare diseases, identify tumor mutations linked to treatment resistance, and find new drug targets.
But clinicians often face a problem: two tumors with the same mutation can respond differently to chemotherapy, and two patients with the same symptoms can carry entirely different DNA alterations.
The reason lies in the complexity of disease regulation. Changes in splicing, transcription, protein expression, and metabolites clearly regulate the cells and their environment, too. Multiomics is the result of integrating these molecular layers, and it has the potential to reshape precision medicine and translational research.
But when information multiplies, so do the challenges. Each additional omics demands more from limited samples and brings vast data that complicates interpretation and puts individual data privacy at risk. This will likely influence the speed of multiomics integration in daily clinical practices.
In this article, we will cover:
• How multiomics speeds up translational research
• The role of multiomics in precision medicine
• The main technical and ethical challenges of integrating multiple data layers
How multiomics accelerates translational research
Turning a discovery into a real clinical application takes time and investment. Multiomics could accelerate this process by providing new and more refined drug targets and improving patient subtyping, lowering the risk of costly late-stage failures.
More informed drug target discoveries
Selecting an accurate target for a drug is crucial to improving success rates in clinical trials. Using multiple molecular layers of information can make this process more accurate and efficient, as it better captures the complexity of human diseases.
This is how Insilico Medicine moved from target identification to preclinical candidate selection in just 18 months. Using their proprietary PandaOmics model, they integrated data on transcriptomics, GWAS, interactome, and scRNA-seq to find new therapeutic targets for idiopathic pulmonary fibrosis.
This led to the identification of TNIK, a serine/threonine kinase that acts as a central regulator of multiple profibrotic and pro‑inflammatory programs, and the rapid development of a small-molecule inhibitor that has already progressed into Phase II clinical trials.
Using multiomics to pick the right patients
Not every patient will benefit equally from the same drug. Multiomics can help stratify patients into more meaningful biological groups, so that early-phase trials are run in the right populations.
This is particularly relevant in cancer, where tumor heterogeneity produces highly variable clinical outcomes. Gastric cancer is a good example, with patients showing resistance to platinum-based chemotherapy. By using multiomics, researchers have identified three distinct subtypes that were previously unknown.
The CS2 subtype showed high stromal infiltration, revealed by combining genomics, epigenomics, transcriptomics, and confirmed with spatial analyses. It had the worst prognosis and the lowest predicted sensitivity to different treatments. This information improves the stratification of gastric cancer patients and supports the enrollment of CS2 individuals in clinical trials targeting stromal-mediated resistance mechanisms.
Beyond one-size-fits-all care with multiomics
Precision medicine aims to adapt diagnosis and treatments to the uniqueness of every patient and disease, increasing success rates and reducing side effects. Multiomics is crucial to bring richer, patient-specific data that informs every clinical step.
Pushing diagnosis past the genome
Whole genome sequencing has proven to be a useful diagnostic tool to find genetic variants causing diseases. But some variants have unclear biological impact, leaving patients without a conclusive diagnosis.
Multiomics can help extend these cases. By looking at the downstream effects of DNA changes, such as changes in transcription and protein expression, candidate variants gain functional context.
This approach supports the reclassification of variants and improves diagnostic resolution in rare diseases. In a national-scale study from the Australian Acute Care Genomics program, integrating whole-genome sequencing with RNA-seq and targeted proteomics increased the diagnostic yield to from 47% to 54% in critically ill children affected by rare diseases.
Multiomics can also speed up precision prevention by capturing a dynamic view of individual risk. Using data from the UK Biobank, researchers developed an AI model that helped predict multiple cardiovascular diseases. Their results showed that combining multiomics with clinical data significantly improved risk prediction up to 15 years before diagnosis and highlighted potential biological mechanisms for treatment.
Informed treatment decisions with multi-layer data
A single clinical condition can hide distinct molecular profiles that will respond differently to a treatment. Multiomics provides useful information to select which treatment to use based on individual traits.
A recent study published in Science examined breast cancer patients who received neoadjuvant therapy. The authors combined epigenomic, proteomic, and clinical data, and found that specific DNA methylation patterns were linked to residual disease after treatment. They developed a machine learning method to predict treatment response that outperformed single-omic approaches.
Clinical implementation is still limited, and more data is needed to support the advantages of multiomics in treatment selection and its feasibility in clinical practice. This is exactly what a study explored in melanoma patients.
The authors found that adding multiomics on top of clinical and NGS data increased informed treatment choices, reaching 75% agreement between multiomics suggestions and clinicians' decisions, and 33% increased compared to NGS plus clinical data. Complex cases benefited the most, showing a significant increase in progression-free survival from 2.03 to 8.08 months.
This was done with a short turnaround time of about four weeks from sample acquisition to final decision. Sample quality was still a challenge, limiting full data integration across technologies.
Why is multiomics still hard to use in the clinic?
The clinical potential is clear, but there are still limitations to using it consistently, reliably, and ethically.
The challenge of clinical samples
Many clinical samples are low-input or degraded, making it difficult to obtain high-quality libraries. This is especially true for FFPE samples, a dominant archival specimen in translational research.
This matters in multiomics, as Brian Kudlow, Chief Scientific Officer at Watchmaker Genomics, explains, “When multiple molecular readouts are needed from the same block, the problem becomes more acute, as every downstream assay competes for compromised input.”
Fragmentomics, an emerging omics based on detecting cell-free DNA patterns, faces similar constraints. It is gaining traction for sensitive tumor detection, but “fragmentomics relies upon subtle changes in fragment endpoints and sizes, making it especially sensitive to preanalytical variables,” says Kudlow.
Making the most of compromised material
Watchmaker Genomics concentrates its efforts on library preparation improvements. Their DNA workflows are designed to “maximize conversion of fragmented, low-input cfDNA or FFPE samples into sequenceable libraries while minimizing added bias and artifacts,” explains Kudlow.
TAPS+ extends this to methylation and genomic variant analysis. Kudlow highlights that “its gentle chemistry preserves native cfDNA fragment structure and supports multimodal liquid biopsy workflows with as little as 1 ng of cfDNA.” This also avoids extra damage and bias of conventional bisulfite methods in already compromised FFPE samples.
Beyond sample quality, batch effects from collection, reagents, instruments, and pipelines add another challenge to multiomics. Kudlow notes that the lack of gold standard reference materials in newer omics makes it hard to reconcile result across methods. He expects rigorous pre-analytical standardization, locked SOPs, shared controls, and harmonized analysis to help push multiomics into the clinic.
Finding the story in the data, and sharing it with care
The complexity of analyzing multiple molecular layers is a real nuance that can risk spurious conclusions. AI models are pushing the field toward more automated and integrative analysis, extracting more information with reduced analytical time.
However, some limitations persist. The lack of training data in rare conditions and cell types can lead to reduced predictive accuracy, and data heterogeneity limits its use in learning models. Other concerns are about human interpretability of black-box models, where explainable artificial intelligence (XAI) can help increase transparency.
In terms of ethical data sharing, the level of individual detail provided by multiomics makes it inherently identifiable. This means the data can be used to match a person across databases, or link anonymous genomic data back to an individual. As Kudlow states: “Ethical concerns can limit cross-institutional data sharing and slow the development of robust clinical benchmarks. Emerging data-sharing infrastructures and privacy-preserving technologies will enable broader collaboration without compromising patient confidentiality.”
That’s why recent work on privacy-preserving federated learning for omics, such as PPML-Omics, is being developed: to balance privacy concerns alongside the data needs required to advance machine learning models.
Multiomics is still early in its clinical journey, but these examples show its potential to achieve more individualized medicine and faster research translation. Their future will be defined by how we integrate these data in a robust, interpretable, and ethical way at scale.
About the author: Nuria M. Wentzien got her Ph.D. from the University of Granada in 2023. After finishing her Ph.D., she discovered that science communication brings together her two favorite worlds: science and writing. Her career reflects her passion for learning and exploring new topics, from a Bachelor's degree in Biochemistry (with a strong focus on human sciences) to a Ph.D. in sustainable agriculture. She brings this same curiosity to every project, making readers feel as inspired by science as she does. Now dedicated to helping life-science companies communicate rigorous science in an engaging way, she is the co-founder of Helixa Communications.