The contradiction of multiomics
SUMMARY
Multiomics is an essential tool in life science research, yet its definition remains elusive and the myriad of methods available has introduced new challenges.
In our 2024 State of Science, we found that multiomics investigations rank in the top five technologies that will make a significant contribution in 2024.
There is an unmet need in moving the utility of multiomics from principle to practice.
This is part one in our series more deeply exploring the trends that we believe will impact life science and health in 2024. Read our summary here.
Multiomics is widely regarded as key to understanding the complexity of biology. But can we align on a singular definition of multiomics? Does this ambiguity hurt or help the field?
Multiomics has emerged as an essential tool in the advancement of life science and health research. Researchers recognize that no single measurement will tell a full story, and to understand disease, develop new therapies, and make more informed treatment decisions, we need a holistic, systems-level view of both patients and other model systems. However, how we accomplish this holistic view has resulted in a bolus of tools and analytics methods that have introduced new challenges.
The term multiomics is so normalized in scientific discourse, that it doesn’t take but a few clicks to find it offered as a service in leading life science companies, making for a crowded market. If you were to ask a scientist what the term multiomics means to them, you’d likely get a response along the lines of, “Some combination of DNA, RNA, proteomics…different classes of biomolecules?” In our 2024 State of Science, multiomics investigations rank in the top five technologies that will make a significant contribution in 2024. The anticipation is there, but what exactly does that mean for the average scientist or patient? Does more data really mean better understanding?
As we continue to realize the evolving landscape of multiomics, there is an unmet need to bridge the gap between its conceptual significance and the practical integration of these multi-dimensional datasets. Today’s tools deliver discrete data sets, which must coalesce during analysis to enable interpretation. Standardization of molecular processing, accounting for biological heterogeneity, and routinely checking for bias in the analysis pipelines are going to be imperative to successfully moving this from principle to practice. Such efforts require collaborative efforts and investment, and in September 2023, the NIH pledged their support through a $50.3 million commitment to “multi-omics”' research on human health and disease. The new research consortium, coined the Multi-Omics for Health and Disease Consortium, is charged with developing strategies for clinical data integration, with a specific focus on ancestrally diverse populations.
In the private sector, companies who embrace rigorous experimental design, robust metadata collection, and work to generate more unified data sources will be poised to win in this space. Collaborations between technology companies and those conducting clinical studies could expedite this, and software tools that leverage AI and ML to make sense of the data will also drive progress (two other themes in the 2024 State of Science). Researchers and reviewers will be looking for compelling validation data, so well-annotated training and testing data sets are imperative.
So, perhaps the enigmatic nature of the term multiomics reflects the complexity of what still needs to be accomplished before the promise is realized. In the interim, technology developers can leverage new terminology to clearly and intuitively differentiate their approach from the generic mess.