: In your opinion, what is the current role of big data in the pharmaceutical industry. How do you see it expanding in the near future?
VR: We know that achieving a step-change in the efficiency, productivity, and quality of biopharmaceutical processes will require significant evolution across a range of technologies and disciplines. One of the most critical areas of need and opportunity is in the field of big data. We collect a vast amount of data on our processes but struggle to translate that data into valuable action in order to understand, improve, and optimize how we make biotherapeutics.
The more I think of the future, I see technologies like big data forcing drug manufacturers and suppliers to revisit their R&D processes, manufacturing process silos, and automation strategies to make them better connected and derive integrated insights. Today, data-enabled discovery & investigation approaches can help unveil information faster, outline multiple scenarios, predict process outcomes and help in implementing robust processes.
There are clearly trends emerging which are focusing on new digital techniques to capture and compute as close to the source as possible, integrating these findings with real-time process analytics and historical information that’s already collected.
: Are there any impediments to using big data in the Pharmaceutical industry? If so, how can they be overcome?
VR: While big data provides us ideas on the potential – as you rightly pointed out – there is a flip side to its implementation as well. Big data speaks to the volume, variety, and velocity of data. I’d then add veracity and value along with the unique challenge that bioprocessing brings with it.
We know the importance of data accuracy in our space. We are dealing with therapeutics and patient outcomes, and the traditional big data tools and models may not be sufficient enough for us to rely on. Before we speak of insights from big data, we need to know how to handle such data, how to organize it, how to preprocess it and then harvest it.
Thus, companies need to be prepared to handle the volume and variety of data, the speed at which it accumulates, the cleaning and preprocessing it, and eventually organizing it the right way to make sense. The industry will have to come together and define the assumptions and models that make this possible.1
: What can industry, regulators, and tech providers do to make the adoption of big data analysis more efficient?
VR: While we are already seeing clear collaboration in the industry, we also need to define use cases that will give more meaning to the efforts. Many companies have invested in data lakes but are struggling to translate them into insights. This demands clarity and focus in purpose. There is a clear need to demonstrate clinical utility to gain acceptance. This purpose-driven approach will drive efficiencies.
Then going beyond – we need to understand how we tie up other initiatives that will be needed to make adoption efficient: what is the data governance strategy, the data privacy and security policies, review of current computing infrastructure, and the most important thing being the right skills that can combine bioprocess domain and digital skills. To be efficient, we need to ensure we approach this more holistically and that we are making the right investments in each of these areas.
: If a pharmaceutical company wants to begin implementation of a big data analysis program to produce better outcomes, are there are any areas/functions that should be looked at first? Why?
VR: Functions such as Manufacturing Science and Technology, Quality Assurance, and Technical Operations would greatly benefit from tools that provide integrated realtime access to process data and insights.
Application of data around process characterization at phase 2/3 has the potential for large benefits as the process will be better understood before moving to commercial scale. This will ensure that the process control strategy, product quality, and yield are optimized before tech transfer.
I also feel we need to start with an approach of creating an analytics “sandbox” (a digital lab of sorts which integrates process technologies with digital tools) and apply that to our processes in an environment that is are less stringent from the regulatory requirements. We are exploring this approach within our own BioContinuumTM Platform Process Innovation and Lab scale application teams.
: Do you foresee the application of big data technology resulting in more collaboration in the industry? Can you explain?
VR: There is a good reason that you hear the cliched sentence that “data is the fuel of the new economy” from various technology and business leaders. In the coming times, data silos will start reducing as industry players come together in the form of consortia or focus groups to define how to work on things like plug & play, standards, and protocols. For example, you will see solutions that allow data to flow easily between functions of discovery and clinical development and then to external partners.
This trend of collaboration is manifesting in multiple initiatives and trends that you see across disparate yet interlinked players:
- The technology players who provide hosting solutions, data management platforms, and new algorithms and languages;
- The regulators who recognize these new platforms and are defining policies that allows the industry to leverage these new methods of data handling & sharing;
- Pharmaceutical manufacturers who are adopting technologies to harness latent information from the product and process data across their value chain;
- And finally, we as suppliers who are integrating our core technologies with tomorrows digital capabilities, making our offerings in-line with the advancements the industry demands.
We all must rely on the strength of our individual expertise & a collaborative approach to deliver on the promise of delivering drugs faster, cheaper, and of the highest quality for our patients.
Resources:
- https://www.cell.com/trends/biotechnology/references/S0167-7799(08)00245-X
- http://glaros.dtc.umn.edu/gkhome/fetch/papers/biopdm08tb.pdf
Reference:
- https://www.cell.com/trends/biotechnology/references/S0167-7799(08)00245-X