Simplifying Analytical Knowledge Transfer in an Externalized World
Industry Insight Jun 09, 2015
The lion’s share of biologics and chemical R&D today is being outsourced to external organizations. As externalization and research virtualization continue to evolve, the task of mining real-time and legacy analytical chemistry datasets will become even more difficult based on deficiencies in the knowledge exchange mechanisms. The deficiencies will become more prevalent as the trend towards bigger and wider partnerships grow, making this a key data management and informatics challenge to watch over the course of the next decade and beyond.
AM: Regarding data management challenges, can we start first by asking how data is shared and communicated today in between a pharmaceutical customer and their CRO partner?
RS: Let me start by saying that of course it depends on the environment, the importance of the data and the relationship the pharma company has with their CRO partner. But in general, the most common way to share analytical datasets generated by partners today is via email, PDF files, Word documents, spreadsheets, etc. A lot of companies will manage their projects on a portal or team platform like Microsoft Sharepoint or simply use an FTP server to share data. However, it seems very few companies have a strongly established mechanism for sharing this data.
AM: As a follow up to that, why is this current collaboration model deficient or even dangerous?
RS: The reality is that scientists don’t need every single piece of analytical data that is generated by their CRO partner. However, when they get to a point where they want the actual data set, it often involves a long process of emailing or calling a representative from the partner company and asking them to provide a copy of the raw data. Sometimes this can take half a day or even a day because of time zone issues or slow response times. Now more than ever, people want their data on-demand.
I don’t know that I would go as far as to say that the current collaboration model is dangerous, but it is certainly risky. Anytime you take activities in support of the research, development, and manufacturing of a pharmaceutical drug, you need to take extra precautions to ensure those processes are carried out in the right way and are backed by correct scientific decisions. I once had a Senior Director from a major pharmaceutical company tell me that at the end of the day they need to ensure that what they are proposing to the market is indeed what they tested in the clinic. They need to also understand, monitor, and control things like metabolites and impurities that are associated with their product. Connecting these things together and hiding behind the scenes of these decisions is analytical data. Some of this work is done in the external network and much of this data remains at the CRO site even after the partnership comes to an end.
AM: Does this collaboration and data sharing model have to be bi-directional. What kind of security risks and considerations need to be evaluated?
RS: In an ideal world, yes, the sharing models should bi-directional to encourage not just data sharing but also research collaboration. However, data pockets also need to be secure and some data is not for everyone’s eyes. This is why things like software multi-tenancy is so important when building a collaboration platform for this purpose. It’s a way to ensure only certain people and certain teams see certain things or have specific privileges.
AM: How important is the cloud to the future of outsourcing and externalization and how relevant is this technology to the IT sectors of Pharma and Biotech today?
RS: I think it is extremely important. It really is the ideal platform for sharing data globally both internally and with externalized networks. A few years ago, I would say it was less relevant because there were still a lot of questions and concerns about data security in the cloud specifically when they are dealing with highlighting confidential and proprietary information. I don’t want to say those concerns have completely dissolved, but certainly the cloud is playing a huge role in IT road maps and strategies in regards to data management and collaboration.
AM:Finally, how does your software and its respective technology address or help facilitate in this challenge?
RS: One thing I did not mention above when describing the process of contacting a CRO representative and asking for data is the form in which that data is available. One of the major issues with data sharing in the analytical space is that data is generated for a wide variety of analytical techniques for an even wider variety of analytical instrument vendors. As an example, one of your CRO partners might not run the same HPLC systems that you do. As a result, if they send you the raw data in their proprietary format you can not open it, much less use it. This example is relevant across all the techniques from NMR to LC-MS to IR and Raman, etc. Our software is designed to remove that burden. Thanks to the strong relationships we’ve built with the instrument vendors in the analytical chemistry space, our software platform enables our customers to take a vendor-agnostic approach to analytical chemistry. Our system can provide users with access to the ‘live’ analytical data for viewing, processing, and analysis from pretty much any vendor format. This really makes it an ideal platform for sharing data with partners that overcomes this traditional limitation of incompatible and inconsistent file formats and software support.
Finally, we don’t just help homogenize and manage this disparate data. We do very scientifically interesting things with it. One of my favorite examples is our Automated Structure Verification (ASV) technology. With this technology we can automatically monitor NMR instruments within the CRO network, pick up the data files, automatically execute some automated processing and analysis routines (including a quality check to ensure the data matches the structure), and deliver the results to a representative from the pharmaceutical company. As I highlighted earlier, it is very important to ensure the quality of data and compounds companies receive from their CROs is high, so I think this is a very good use that many of our customers are excited about.