Generating and protecting knowledge with electronic laboratory notebooks

In the face of spiralling R&D costs, currently estimated at $800 million to produce a new drug, the discovery industry is constantly searching for ways to keep expenditure low, increase productivity and meet ever-stricter guidelines imposed by regulatory bodies.

One valuable strategy for drug discovery is to examine how maximum value can be extracted from a life science organization's primary asset — knowledge. Exploiting and manipulating this knowledge effectively is crucial in the competitive environment of creating new drugs, yet many organizations still rely on the traditional medium of paper-based lab notebooks to record and store their valuable experimental research.

The inherent disadvantages of storing experimental information in this way include difficulty in:

• Sharing data (there is probably only one notebook where the data is stored).

• Accessing data (locating the exact page containing the required data is time-consuming).

• Making changes (if, for example, experimental conditions change during the course of the experiment).

• Extracting information to create reports.

• Querying the data to find the exact information you need in the context of the experiment.

• Proving compliance to a process or regulatory set of guidelines.

While paper-based lab notebooks may have served the industry well in the past, there is a growing demand to bring the industry into the 21st century and adopt a more efficient method of recording, accessing and searching experimental data and corporate knowledge.

The benefits offered by ELNs

Electronic laboratory notebooks (ELNs) are electronic record-keeping tools that provide an IT solution to the problems presented by storing vital research information on paper. All of the knowledge associated with an experiment is stored securely in a central location, where it can be searched and accessed with comparative ease.

Locating and extracting the exact data that a researcher or manager wants, when they want it, simplifies report creation and improves communication across an organization. This ensures decisions are based on accurate and up-to-date data.

In particular, the new generation ELNs solve the following issues facing drug discovery organizations:

• Intellectual property (IP) protection — life sciences organizations invest considerable time and money on protecting IP because of the limitations of paper lab book systems.

• Compliance/QA/QC — meeting stricter industry compliancy regulations using paper lab notebooks leads to time-consuming and inefficient paper chases while tracking and authenticating amendments.

• Research efficiency — the significant amount of time that scientists spend on writing and administering reports from paper-based experimental results could be used more efficiently to conduct research.

IP protection

Problems. Corporate knowledge and IP are the fundamental building blocks of any scientific organization, but if a company can't easily access its own data, and so is unaware of the information it possesses, protecting IP can seem an impossible task.

Paper-based information poses many problems when proving, for example, 'first to invent' in a patent case or legal proceedings. Information is not necessarily chronological or legible. The context of associated experimental information, such as graphs, data spreadsheets or results printouts attached to a page, can easily be mislaid.

Solutions. An electronic version of a company's lab books allows efficient knowledge mining because of the ability to search for, locate and extract information easily. An ELN can significantly reduce the time and cost of searching for previous data and knowledge. Improved IP collection, storage, protection and measurement fulfil essential scientific legal requirements. The risk of lost IP when personnel leave the organization is reduced, making adherence to legal compliance issues easier to prove.

Compliance

Problems. Paper notebooks lack an audit trail to verify what was entered when and by whom, and time and effort is wasted checking for signatures, date stamping and sign-offs that authenticate data entry and track amendments. With no easily visible history of the experiment and no in-built security to control access to the experimental data, an organization faces challenges when trying to enforce good manufacturing practice

(GMP) and good laboratory practice (GLP), and to adhere to regulatory guidelines such as 21 CFR part 11.

Solutions. The best ELNs can generate audit trail information, so that each addition and change of content is tracked, time-stamped and authorized in a manner that is fully compliant with 21 CFR part 11, demonstrating a clear advantage compared with recording experimental data using paper lab notebooks.

An ELN can typically provide improved traceability and a stronger enforcement of policy and procedure by providing control over the data entry process to enforce GMP and GLP.

Privileges associated with user roles and groups can control access to system functions and visibility of data, while electronic time-stamping and author sign-off provides an audit trail of all access to and changes to the data.

Research efficiency

Problems. Researchers need a tool that supports research and discovery by removing some of the administrative and practical headaches that plague science today (Figure 1).

Figure 1 An ELN securely stores and manages data, allowing scientists to concentrate on research.

It is currently estimated that a researcher spends anything from 5–25% of available research time formatting, cutting and pasting, and transcribing data into paper notebooks, depending on the level of regulation in a laboratory. Writing a report, for example, often involves copying from a paper lab book into a spreadsheet or reporting application, duplicating the task of data entry. The chronological nature of a lab book does not lend itself to quick and efficient searching — experiments can be split over different pages or even books.

Solutions. Studies show that an average scientist will save 20% of their time through use of a new generation ELN. At one day a week for each scientist, the estimated time saved for a typical laboratory team of 40 researchers equates to 160 days a month, or 384 weeks a year, of human resource that can be used to increase efficiency and productivity elsewhere.

Significantly faster retrieval of data from past experiments using a good ELN can free up an organization's resources to extend its research and discovery beyond the limits imposed by the administrative overheads and duplicate activities inherent in paper-based notebooks.

Ease of use across disciplines

Problems. Previously, ELNs have catered for individual chemistry applications and did not provide a working environment that was easy to use across multiple scientific disciplines. For an ELN solution to be successful and meet the range of scientific needs, it needs to be able to cater for very process-driven input from departments such as analytical chemistry as well as less process-driven groups such as exploratory biology.

Solutions

An IT solution should provide a working environment that is agreeable and pleasing to the researcher, without forcing the scientist to learn and adopt a totally new way of working.

An ELN with an open standards data format — that can store and search information of any data type — would be flexible enough to meet the needs of scientists across the scope of disciplines, from molecular biology to analytical chemistry, and pharmacology to genomics and proteomics, whether in academic research laboratories or a highly regulated GLP environment (Figure 2). An ELN with such an information infrastructure would accommodate easy integration with other systems and the scalability to add new functions as 'extensions' for individual scientific areas as and when needed.

Figure 2 A successful ELN solution accommodates multiple scientific disciplines.

One way an 'open standard' architecture can be achieved is by storing record data as finite objects rather than in standard document formats.

Recording data using component objects can:

• Enhance searching — getting straight to the data.

• Make reporting easier— rather than copying and pasting into reports, objects can be used instead, which retain a link to the original information, making maintenance of information within a report easier to manage.

• Business process management and business decision support.

The use of objects means that individual packets of data can be electronically signed and also used elsewhere in the organization through the use of business process management (BPM) and business decision support (BDS) tools — a feature that is becoming more and more appropriate in an enterprise ELN solution.

The perfect ELN solution

The industry requires a software tool that emulates a traditional paper lab book by collecting all relevant information concerning experiments, but also provides a safe repository for all experimental data. The need is NOT for a word processing package, a spreadsheet package, a chemical structure editor or statistical package, which already exist and provide useful but specific functions. The data generated by researchers is often scattered around a computer hard drive and other corporate databases and laboratory information management systems (LIMS).

The perfect ELN system captures all the information that a scientist generates and stores it in a manner that allows easy but controlled access, focused searching, data mining and reporting — functions that a paper lab book cannot provide. It also permits additional experimental information to be added as it becomes available whilst a tamper-proof tracking system maintains an audit trail of all the changes made.

By providing a central location to store all the information and data from an experiment, an ELN solution allows know how to be disseminated through an organization easily as new procedures are implemented centrally and viewed immediately by users.

An ELN should store and search any data type, from plain text, images, sketches, scanned documents, and Microsoft Office documents, spreadsheets and presentations, to chemical structures, protein sequences and structures and gene sequences. In addition, linking to other data in secure repositories is possible, as well as access to web site URLs and PDF documents.

Any systems already used in the laboratory should be able to seamlessly integrate with a platform- and format-independent ELN, so that any database can be 'plugged in'. For example, this allows a researcher to go to a LIMS systems, pull out the required image file, data table or binary output, and incorporate it into an experimental report, maintaining the context of an experiment.

The future of ELNs

The ELN market is changing rapidly as the industry understands more and more about what it requires and what users expect from e-systems.

Some expectations require that an ELN should stretch out across all aspects of the drug discovery and development process, from genomics and proteomics, underpinning target identification and validation, all the way through to new chemical entity (NCE) launch.

At present ELN solutions are limited to capturing data in a workbook management type of function, while others believe there is scope for the ELN to expand beyond results data management to an entire data management system for a complete organization.

There is evidence that other industry sectors, such as aerospace, that also face a significant scientific data management challenge, have perhaps been quicker to understand the value of a more integrated approach to data management. Standards and technologies have evolved to support collaborative data management which spans not just functional groups within an organization, but entire communities across globally dispersed supply chains.

Given the growing importance of contract research organizations (CROs) in the drug discovery process, perhaps it is time to learn some lessons from other sectors. Respected academic bodies such as the Warwick Manufacturing Group are already being asked to help apply 'lean' design processes from the aerospace world to the healthcare sector. Perhaps the next major target for the ELN will not be further improvements in research productivity within therapeutic groups, but rapid and secure information flow between communities of research partners — a 'supply chain' for IP as opposed to physical goods.

What is clear in the short-term is that as requirements for e-laboratory systems grow, they are inherently required to link to other data sources, instruments, catalogues and databases.

To capture data, knowledge and IP in an open-standards-based architecture and open data format will undoubtedly help the retention and validation of information in the future as it will be easily retrievable and maintainable. ELN records will be required to last for many years and for at least as long as a drug is on the market. Based on the example of aspirin‚ which was discovered in the late 1800s, the time span that data would need to be kept for in an accessible format would be more than 100 years.

The storage of paper records for this length of time — or longer — poses problems. In contrast, digital media can be backed up and archived very cost effectively to ensure that the information is preserved. E-records offer the overwhelming bonus of storing multiple experiments and their associated data for the entire lifetime of a drug, in a format that can be searched, moved, copied, protected and archived with comparative ease.

Paul Denny-Gouldson is product manager at IDBS, UK.

References

1. J.A. DiMasi, R.W. Hansen and H.G. Grabowski, J. Health. Econ. 22(1), 151–185 (2003).

2. R. Lysakowski, Comparing Paper and Electronic Laboratory Notebooks — Part 2, Scientific Computing and Automation, May, 1997. www.censa.org

3. Achieving Success with Electronic Laboratory Notebooks, ELN Deployment Case Study, presented at General Electronic Laboratory Notebooks Summit, IQPC, September 2005, London, by David Dorsett, VP and General Manager of Software, Symyx IntelliChem Inc.