January 14, 2019
by Sanji Bhal, Director, Marketing & Communications, ACD/Labs
I truly was not just looking to be provocative as I sat down to write this post…the term ‘necessary evil’ has stayed with me since I first heard it used by a pharma/biotech executive, who has since become an ACD/Labs customer, outspoken in his belief that analytical data sharing and management merits the necessary effort and investment.
The discussion began around the investment R&D organizations make in analytical chemistry—capital investments in instruments and hardware totaling multi-millions; expenditure on consumables such as solvents and chromatographic columns which are essential for the analysts employed to maintain and run the instrumentation, process and interpret results, and help provide the answers sought.
At ACD/Labs we have assembled an extraordinary amount of experience in helping scientists and their organizations get maximum value from their analytical data. Not only by helping extract answers from data efficiently—through application of chemically intelligent data analysis algorithms, industry-leading structure verification tools, and automated processing and analysis allowing ‘review-by-exception’; but also enabling customers to store the knowledge gained in a scientifically intuitive manner with the context of the original experiment to maximize its value in future review and re-use.
Analytical data is necessary
The sentiment in the title does not take away from the fact that almost anyone you speak with about the topic will agree that analytical data is a necessity in R&D. Knowledge gained from analytical experiments allow scientists to:
- Determine and confirm the chemical identity of pure substances and the identities of complex mixture components
- Measure substance purity and/or composition
- Determine causes of process and manufacturing problems
- Determine optimal process parameters
- Prepare necessary proof information for regulatory bodies
Things begin to get a little hairy when you consider all the different types of experiments necessary to draw definitive conclusions. Rarely will a single analytical technique enable a definitive conclusion of the identity or composition of a substance—a variety of experiments are required to provide justifiable confidence—perhaps LC/MS and NMR at minimum. Now, think about the number of vendors from whom those instruments might be acquired, the number of formats the resulting data will be in and an ugly picture begins to be revealed. The warm, fuzzy feeling of how important analytical chemistry is begins to be overridden by a panicky feeling of ‘how do we manage it all?’
What Housekeeping and Analytical Data Have in Common
As the neat freak in my home, I liken analytical data management to housekeeping. It’s much easier to maintain a neat and tidy household if you pick up those dirty socks and return things to their rightful places as you go along than to clean up a ‘pig-sty’. In the same vein, if/when producing a variety and quantity of analytical data, one should follow well-defined processes and procedures for handling it and use software systems that are designed to manage it effectively. The biggest difference between these scenarios is that cleaning my house gives me instant gratification because I ‘see’ neat and tidy, whereas an organization implementing a new analytical data management strategy will need to devote some effort before they realize gains in ROI (though their data will immediately be better organized and accessible).
Effective Analytical Data Management is Worth it!
We understand that it can be difficult to motivate scientists to get excited about managing their data better, especially when those putting in the greatest effort may feel they have the least to gain in the short term. How many times have we heard ‘I know where the data is and how to find it’ from the analyst? What they forget in that instance is the number of times they have to answer the same questions—”Bob can you send me the LC/MS data for x?”. “Please can you expand this section of the spectrum and send it back to me? I think there’s something under there”. “Bob can you summarize the characterization of the following process route impurities for the team meeting tomorrow?” Poor Bob forgets that he spends so long fulfilling requests for experiments he’s already run that he scrambles to fit in all the analyses still awaiting his attention.
It really does benefit everyone to be able to search and access the data that is the source of so many critical decisions. Not only that, once that data is easily accessible more people in the organization benefit from the value it provides and those that described analytical data as a ‘necessary evil’ begin to see the light. In all honesty, many ACD/Labs customers that store and manage large quantities and a variety of analytical data automate sweeping of data directly from instruments, followed by processing, analysis, and storage into a centralized database. The processed data can be ‘reviewed by exception’. Automation reduces the burden of data management on scientists and enables the organization to get maximum value from their investment in analytical data collection.
How is analytical data viewed in your organization, or are you scared to ask?