Functionomics: operationalising big data for personalised rehabilitation care 

Functionomics holds promise in rehabilitation medicine, which operates from a biopsychosocial paradigm, focusing on improving the daily functioning of individuals. Currently, 2.4 billion people live with a health condition that would benefit from rehabilitation globally. To put the potential of this data into perspective,if each of these patients generated 10 megabytes of functionomics data, the volume of functionomics data would amount to ~24 petabytes — the equivalent to 8 million HD movies. Within the PREPARE project, we anticipate using data from approximately 300,000 patients, totalling many ‘bytes’ of functionomics data. One can imagine the potential in transforming this large amount of data into actionable insights for personalising daily rehabilitation practice.  

​​​​​How big is one petabyte? 

Personalised medicine is driven by the analysis of big data, commonly referred to as ‘omics’ research. Genomics, transcriptomics, proteomics, metabolomics, and radiomics have already shown great potential for revolutionising personalised medicine. Functionomics introduces a new dimension— the study of high-throughput data on daily functioning associated with health. Critical to the success of all ‘omics’ initiatives is the concept of machine actionable data. This means that in order to unlock the full potential of functionomics data, we need to deal with the challenges posed by unstructured and inaccessible functionomics data in rehabilitation medicine and related disciplines. 

The guiding framework that can help us to unlock the potential of functionomics are the The FAIR principles—Findable, Accessible, Interoperable, and Reusable. The FAIR data principles are embraced by the international community, like the G20 and European Open Science Cloud, to promote appropriate data (re)use. By making data findable, accessible, interoperable, and reusable, functionomics can drive new knowledge discovery. Recently, my study was published on how functionomics can be realized through the FAIR data principles, leveraging the ICF. This publication paves the way for further development of functionomics in rehabilition care.  

‘omics’ research initiatives 

My passion for data analysis and data science started during my PhD on prehabiliation in spine surgery. I enjoyed  analysing data; it served as the means to unraveling the meaning behind said data and addressing my research questions—the gateway to new knowledge. At the same time, I spent hours and hours on the tedious task of manually inputting paper-based data into datasets and gathering data from different sources, such as hospitals and physiotherapy practices. This is a common challenge in rehabilitation related data. I was also troubled by being the sole custodian of my research data, fearing that it would be lost after my PhD. I understood that my data held a wealth of untapped knowledge, if only more people would be able to find, understand and use it. This is when I recognised the immense potential of making ‘functionomics’ data FAIR. It goes beyond enhancing accessibility and interoperability; it’s about democratising knowledge and enabling researchers and practitioners to unlock insights and address important research questions together. By connecting data silos and facilitating data reuse, we can stimulate innovation, drive evidence-based practice, and ultimately improve patient outcomes in rehabilitation care. 

This is why in the discipline of rehabilitation it is important to embrace the FAIR principles and invest in these proposed solutions, making our data FAIR. An example of this is investing in developing common data models that are able to capture the complex intricacies of rehabilitation treatment and outcomes and stimulate international data capturing and mapping protocols in rehabilitation research. In PREPARE, we will showcase the transformative power of functionomics through machine actionable rehabilitation data. To operationalise the FAIR principles, we rely on the application of OHDSI tools. However, it’s important to note that while these tools are often designed for biomedical data, they may not always work in making functionomics data actionable. Nonetheless, within the PREPARE project, we aim to develop innovative solutions that will bridge this gap, in turn making functionomics actionable. 

Together, we invest in a future where functionomics transforms data into actionable insights, reshaping the landscape of personalised rehabilitation care. One way of supporting this journey is by joining our OHDSI rehabilitation working ​​group. Here, we collaborate to make observational rehabilitation data FAIR and collectively address common challenges in rehabilitation research. In this community, we share a mutual passion for maximising the potential of rehabilitation data to optimise rehabilitation care. 

Author: Esther Janssen, Radboud University