The Computational Data Paradigm and The Rise of the Bioplatform

Over the past six months, the COVID-19 pandemic has thrust the MedTech industry into the spotlight. Investors, consumers, and the general public have started to become acutely aware of how crucial human health is not only in economic terms but also within the applicable realm of social constituencies. The MedTech sector is unique because it is continuously evolving with novel scientific advancements and innovations to tackle some of the most pressing health issues in our world today. However, only within the past decade have we seen the rapid evolution of harnessing data to unlock novel treatment paradigms and accelerate scientific breakthroughs. Healthcare and experimental data used to be arbitrary constructs, with vast amounts of data libraries procured and stored with no efficient way to extract insights effectively. We refer to this data form as “stagnant” data, whereby the data could only convey detailed or basic information that could not be extrapolated and analyzed to unleash powerful health-related insights. We find ourselves in a new era where this once stagnant data can now become a living interpolation conveyed through the emergence of cloud computing, artificial intelligence (AI), and machine learning (ML) algorithmic data tools.. Critical insights and information that have been locked in data can now be translated and adaptively analyzed to accelerate advancements in a plethora of MedTech applications spanning digital health and clinical decision support to drug development and discovery.

This acceleration of data interpretation has led to the emergence of “bioplatform” companies that combine proprietary algorithmic software, intellectual property, and a robust clinical/biological database. These bioplatform companies will allow for the accelerated construction of products and services paired with the symbiotic ability to partner with healthcare payers, providers, and pharmaceutical companies that can utilize the platform to develop their products and services faster, which ultimately lowers development costs.

The emergence of bioplatform companies has shown immense promise in the biopharmaceutical development space. Drug development has long been noted as a slow and arduous process, often taking a decade and billions of dollars of sunk R&D costs to develop a drug and hit the market. The benchmark standard of pharmaceutical development has always been the in-vivo animal model, which has been plagued by issues of translational efficacy to human trials. What if we started with human data? This distinct notion of using human data to treat human conditions is not proprietary; however, there has been limited progress on the capacity to analyze human data and translate it into clinically significant results.

This is where the convergence of computational approaches comes into play. These approaches seek not only to optimize drug molecule discovery, targeting mechanics, and delivery systems, but revolutionize the way we go about translating laboratory research into functionally significant clinical outcomes. Target-specific machine learning can be integrated into high-throughput sequencing (HTS) and site identification by ligand competitive saturation (SILCS) to optimize predictive binding capacities. Utilizing computational approaches in single-cell multiomics can lead to a better understanding of individual cellular dynamics and rendering 3-dimensional disease models that can mimic human disease pathologies ex-vivo. In conjunction with advancements in 3-dimensional computerized axial tomography bioprinting technologies that can now print live human cells with 95% viability, we can now test and screen drugs ex-vivo to gauge how these drugs and therapies will react in clinical trial settings. A massive paradigm shift in R&D processes will occur when we understand how a treatment or therapy will react with human data before clinical trials. These computational and integrated data approaches further extend beyond drug development to clinical trial management, in which the “digital twinning” of clinical trial subjects has become possible. Clinical subject data can create virtual phenotypic twins of actual patient subjects without the need for real patients to be physically present. Computational approaches in applications for oncology and radiology have become significantly more efficacious. Specifically, analysis of patient data has led to more precision-based combinatorial cancer treatments as well as more refined disease prognosis from analysis of patient CT, X-Ray, and MRIs.

The potential of harnessing and analyzing data has become vastly accelerated with the use of computational approaches, however, when it comes to AI/ML, the analysis algorithm is only as good as the underlying data on which it trains. Thus, training algorithms with human data allows for the more accurate interpretation and production of clinically significant outcomes. Getting access to high quality tagged and labeled data is a current bottleneck that we will have to overcome in further optimizing the full utility value of computational approaches. When it comes to developing a MedTech product, the more data-centric the method, the better the scope of treatment will be. The emergence of bioplatforms, in conjunction with computational approaches holds immense promise in unlocking novel approaches to treating the most critical human health conditions.