The Practice of Analytics in Life Sciences
In my previous blog, I mapped out what I call the Analytics Continuum in Life Sciences. I demonstrated how each type of analytics, from descriptive to predictive to prescriptive, requires a different learning curve and set of tools to gain the most ground in terms of research and progress.
With the application of advanced analytics and big data technologies being a hot topic across all industries right now, I thought it would be useful to actually map out how scientists (data and otherwise) are deploying the actual practice of analytics in Life Sciences, and how this relates specifically to the drug discovery process. Let’s get started!
Where Analytics Come into Play in Drug Discovery
Today, analytics techniques have been adopted across all phases of the drug discovery process, which can be coarsely grouped into (1) Research and Development; (2) Manufacturing; and (3) Commercial (inclusive of post-market surveillance). These phases can each be mapped to certain long-standing analytics activities, such as:
- Data mining for new potential targets (R&D)
- Analysis of large high throughput screening (biological testing) datasets in order to identify new potential drugs (R&D)
- Optimization of the biological activity of a potential new drug through statistical/mechanical model-driven changes to the structure or formulation (R&D)
- Simulation of the clinical effectiveness and safety of a potential new drug (R&D)
- Optimization of the manufacturing process through simulations (Manufacturing)
- Determining when, where, to whom and at what price to sell a drug, etc. (Commercial)
It’s All About the Data
As you can see above, analytics play a huge role in every phase of drug discovery, and the effectiveness of employing predictive and other types of analytics techniques is of course predicated upon the quality of the data, meaning that it must be reliable, clean and valid to render actionable results. But the infrastructure behind the data can be just as important as its quality.
Along with investments made in the analytics infrastructures within (and outside of) pharma, companies have been making similar scale investments in the ability to ingest, manage and deliver data within their technology stacks. Indeed, the size of these investments has led many companies to pursue common (hosted, multi-tenant, pre-competitive) alternatives to managing the large volumes and varieties of data necessary to execute effective drug discovery programs.
The Future of Analytics
Analytics platforms have not yet reached the level of maturity needed to act as a standalone enterprise solution and are still seen as competitive assets within most pharmaceutical companies. As such, a wide variety of both internal and external resources are responsible for deploying the techniques described above, and it could be years before analytics are as ubiquitous as something like Enterprise Resource Planning using platforms like SAP ERP are today.
As the increase in the volume of data provided through new chemical, biological, clinical, process, epidemiological, economic and other channels continues to grow, a new era of analytics is emerging with greater demand for analytics compute services and/or faster algorithms. The analytics service provider landscape external to the pharmaceutical companies is diverse (often coming in from other business verticals) and will continue growing until pharma itself catches up to other industries.
Join me next time as I discuss the promise of analytics in life sciences, and what goals analytics are supposed to achieve for those who employ them.