Editors' Note: Due to an editorial error, the LIMS story that ran in print was a preliminary version. The following is an updated and complete version.
Feb. 12, 2007 | Until recently, laboratory information management systems (LIMS) were largely confined to providing mundane data management — just keeping track of numbers, more or less. Today, though, scientists expect much more from their LIMS, including the connection of virtually all lab instruments, automated data collection, standardized data output, tracking workflow, and more. Such requirements are propelling LIMS from mere number-trackers to knowledge-creators.
LIMS can generally be divided into two categories: generic and specific. As John Bainbridge, general manager of LIMS at Applied Biosystems, says, “Generic LIMS apply to multiple market segments, and specific ones serve a particular market or workflow.” Applied Biosystems offers a generic LIMS platform, but specific products make up a relatively new approach.
Thermo Fisher Scientific (TFS) is one of the biggest providers of purpose-built LIMS. The company markets four solutions for pharmaceutical applications, ranging from discovery to bioanalysis to manufacturing. Dave Leitham, director of informatics products, says, “LIMS are moving from generic one-size-fits-all to purpose-built applications to reduce the risk associated with customizing the system to meet your specific needs.” Despite being specific, LIMS let users make modifications. For example, TFS’s Darwin LIMS serves analytical R&D and quality control for pharmaceutical manufacturing, and helps set up testing by asking what is being made, say tablet or injectible. Dave Champagne, vice president and general manager of the informatic business, says, “The functionality found in Darwin complements our other pharma offerings to meet the specific needs found at each stage of the drug-development pipeline.” He adds, “Our Watson product is used by virtually every top pharma company in their bioanalytical labs because it is built expressly for that type of work.”
GenoLogics Life Sciences Software is also introducing a science-specific LIMS for genomics called Geneus. Tanis MacSween, marketing communications manager, says, “Geneus is purposed for gene expression and genotyping, and lab and scientific data management.” This company already offers LIMS for proteomics, and MacSween says that they will continue to expand to other life-science areas.
Ocimum Biosolutions, a fast-growing Hyderabad (India) bio-IT company, offers three specific LIMS: Pharmatracker, a multi-platform system for R&D and manufacturing applications; Biotracker for genomics, proteomics, and biobanking applications; and Toxchek, a pre-clinical LIMS for managing general toxicology data.
Still, some experts don’t believe in specific LIMS. Steven Neri, director of pharmaceutical business development at LabWare, which has been providing LIMS software for nearly 20 years, says, “There is no such thing as a purpose-built application. Product enhancements conceived in one market space, if applied correctly, can be implemented to solve problems and provide benefits that were never even contemplated when the enhancements were originally created.” Moreover, what LIMS are asked to do keeps changing. “Now LIMS is taking a more proactive role,” Neri says. “For example, a LIMS might prompt a person — based upon key events or business conditions — to find a sample and do specific work on it. It can automatically add a test, being intelligently controlled to run processes based on rules.”
To find a balance between generic and specific LIMS, many vendors offer customers a menu of modules. For instance, LabVantage Solutions’ Sapphire LIMS uses modules for research & development, quality management, biobanking, and so on, offering clients the advantage of managing one content-rich LIMS across the enterprise.
Other companies also approach LIMS from a put-together-the-pieces strategy. Accelerated Technology Laboratories’ Sample Master LIMS can be outfitted with ten modules, including sample tracking, data entry, and electronic data transfer. Core Informatics offers customers a selection of components or will customize them. LabWare’s LIMS software also includes over 100 functional modules. Modular approaches can also suit smaller companies. The MSC-LIMS can start off with a single user and grow as needs expand.
Overall, LIMS take on continually more tasks. David Bearss, chief scientist at SuperGen, a drug-development company, points out that LIMS can provide analysis and visualization as well as collect and store data. “Ideally,” he says, “its an end-to-end system that runs the whole process.”
When asked about the most interesting recent advances in LIMS, TFS’s Champagne says, “Open systems and a modular approach help with interoperability in the lab.” He adds, “We are building new applications with Microsoft’s .NET platform, which provides cutting-edge presentation tools that cannot be matched by older technologies. This platform makes it easier for customers to configure, update, and maintain software and connect it to other parts of the lab.”
At LabWare, LIMS started with integration in mind. One aspect of integration is instrument interfacing, and LabWare’s LIMS supports a library of hundreds of off-the shelf interfaces to instruments. Another aspect of integration is platform flexibility. LabWare’s founder and president, Vance Kershner, decided that his company’s products should be written in an object-oriented programming language that runs in any Windows environment. Moreover, LabWare’s LIMS does not rely on functionality from any particular database, so it can use a variety of them: Oracle, SQL Server, Access, and so on. Neri adds, “The database can be running on a UNIX server, a Windows server, or even a mainframe.”
Rather than worry about language, GraphLogic tries to bypass it. Its LIMS depends on an object-process graph, which makes up a computing architecture that needs no other computing language or database. The nodes and edges of an object-process graph store information about the program, and it can be modified by changing those nodes or edges.
But language issues in LIMS also arise from ordinary words. As companies expand worldwide, LIMS need multilingual capabilities. PerkinElmer’s Labwork ES LIMS works in Chinese, English, French, German, and Japanese. LabVantage’s Sapphire also works with a host of languages including Chinese, Japanese, Arabic, Swedish, German, Spanish, and French on the same enterprise LIMS. “An organization would need to run multiple application servers or clients around the world with a thick-client system (data processing locally) to achieve this same level of internationalization,” states Ronald Kasner, LabVantage’s vice president of corporate development.
Making LIMS compatible with many databases and languages, however, does not guarantee easy integration of data. An organization might have information scattered across multiple databases, but want it all on one. Mark Taras, of Valiance Partners, which produces software and solutions for a range of information-related problems, says, “If you have three or four systems in a lab, and want to migrate it all to one LIMS and make them talk to each other and validate it all, it could be a nightmare, especially the testing and validation required to qualify it — but we have software that does that.” Still, that software needs to be fine-tuned for every migration.
Other LIMS companies — such as Waters Corporation, which markets NuGenesis SDMS — also see the ongoing need to improve integration. “The biggest challenge with many of our customers,” says Pat Martell, director of informatics product marketing at Waters, “is the integration between different applications that they have.” He adds, LIMS is a piece of this complex mosaic that is changing, and it is evolving to help solve the analytical laboratory’s work efficiency challenges.”
Simon Wood of STARLIMS agrees. He says. “Web-based systems are easy to install and support. There is no software on the client’s device.” Web-based LIMS can even improve customer relations in many cases. “For a quality-control lab,” says Wood, “a customer can log on to check in on the progress of a job. As a result, the organization running the QC lab doesn’t need to spend time fielding phone calls about where a sample is.”
Keeping scientists on track also demands great service from LIMS vendors, and they know that. In looking at the trends toward globalization and standardization, Bainbridge of Applied Biosystems says, “This will require not only a great technical solution, but the services capability behind the solution to enable the customer to effectively roll out enterprise software.” Bainbridge believes that a LIMS vendor can help customers balance the needs of locals sites within a global organization.
For many biopharma companies, the value of LIMS still centers on numbers. “The amount of information that we collect is exponentially more than it was ten years ago,” says David Bearss, chief scientist at SuperGen, “and it grows every year.”
Bearss and his colleagues build computational models to make predictions about potential drugs that can be tested in the lab. For example, they built a discovery engine called CLIMB (ComputationaL Identification Modeling Biology). In essence, this engine drives the workflow for each process in drug discovery and development. “It automatically provides a status report of each process,” says Bearss, “and that is available to team leaders through a web-based management application.” This will provide SuperGen’s leaders with something that they’ve long wanted — giving project managers the ability to put new processes in the pipeline and reallocate resources on the fly. With CLIMB, project managers can monitor which scientists are working on specific projects, what animals and reagents are needed, and so on.
Often, a perfect LIMS solution is not available on the market. To build its required LIMS capabilities, SuperGen relies on off-the-shelf products as well as in-house applications. “We use lots of third-party offerings that we have integrated into our applications and our workflow,” says Bearss. To make such a complete system work, SuperGen hired software experts and then taught them the science.
Many tasks require software engineers and research scientists working together. Therapeutic-treatment developer Lexicon Genetics has an ambitious project called Genome 5000 — knocking out 5,000 mouse genes in five years, to identify the best pharmaceutical targets in the genome. This involves keeping track of 300,000 mice each year, many of which undergo more than 120 experiments apiece.
To handle so many experiments and so much data, Lexicon began by exploring the available software. “There was none,” says Christophe Person, Lexicon’s vice president of informatics. “We wanted software to track every business process, from lab to animal facility, breeding, genotyping, physiology experiments.” To get that, Person’s team built its own LIMS. “When people come to work, our LIMS provides a list of the experiments that should be run that day,” says Buckley Kohlhauff, Lexicon’s senior director of informatics. Although this system uses some commercial software, it is largely proprietary code. In such large and custom projects, companies often need to develop the LIMS themselves.
Nevertheless, Lexicon’s custom LIMS still has a few shortcomings. “It would be great to have a knowledge-management system where scientists could drag and drop pieces of information and design the report that they need at their level of management. That is not quite there yet,” says Person. Part of the problem comes from security, especially in the pharmaceutical business. “Pharma tends to work on a need-to-know basis,” Person says.
Some LIMS address such concerns. LabVantage’s Sapphire includes role-based security, and its Evergreen configuration tool, as Kasner explains, “offers organizations a much greater level of configurability than other LIMS systems, so that customers can tailor the solution to meet their specific needs, rather than adjust their operating procedures to fit the LIMS.”
More than Numbers
Today, LIMS go far beyond numbers. STARLIMS, for instance, forged an agreement to integrate the EMC Documentum document-management system with LIMS. “This way,” says Wood, “scientists can manage documents produced in the lab and the issues of knowledge management.” He adds, “At the moment, LIMS manages highly structured data that allow standard-database searches. Adding documents provides more of an opportunity to look at knowledge in the lab.” This can include anything from following threads of information created in e-mail messages to searching for every document that includes the name of a compound, gene, or disease. “Then,” says Wood, “scientists can look at this information for similarities or even ask the LIMS to find the similarities.”
Another growing application is biobanking. LabVantage’s Sapphire BioBanking Solution, “addresses the unique challenges of specimen collection and banking for pharmaceutical discovery and clinical operations, academic and biosciences research centers, medical institutions, and contract research organizations,” says Kasner. For example, the Multiple Myeloma Research Consortium is using Sapphire to manage samples collected from around the world for genomic analysis, pre-clinical validation efforts, and critical correlative analyses.
Still, this trend is only beginning. “LIMS on the market are usually interested in the final results from instruments,” says Wolfgang Lemmerz, senior solution architect at Waters, “ but you need supporting information so that you can go back to the raw data. This is what the market is looking for: a complete view.” To get that, LIMS must track everything about the experiments performed on a gene, cell line, or compound, and add that to any related documents. Bundling all of this information with the results — and even idle thoughts about the work — turns into true knowledge management. As Lemmerz says, “Knowledge management requires that you get together as much information as possible and make sense of it. We need the science behind it.”
Documents can also be tracked in fields outside of molecular research. For example, Applied Biosystems recently released its Forensic Navigator, which manages information related to case folders, crime scenes, and analytical results. “The system provides access security with audit-trail capability,” says Bainbridge. “It tracks where and when samples have been stored as well as the results of investigations — including DNA genotyping, toxicology, ballistics, trace analysis, etcetera.”
Despite all of the talk about integration and data standardization, these issues still cause problems. “One of the things that we have the most trouble with is standardization of data sources and formats,” says Bearss of SuperGen. “It’s still hard to integrate instruments in the workflow process, because they all use different formats.” Consequently, many LIMS users must generate custom applications that extract data and translate that into a common format.
Nonetheless, putting everything in the same format — especially an open one — might not happen soon. Instrument makers must protect their market space, and using proprietary code enables that in many instances. “With scientific instrumentation,” says Bearss, “it’s not a huge market. It’s pretty competitive, which makes it tough on the vendor’s side of the game.” Nonetheless, Bearss offers one potential solution. He says, “The processes can stay proprietary inside the instrument, and then dump the results as XML.”
To enhance integration, some LIMS already use open architecture. MacSween of GenoLogics, for example, points out that “Geneus and Proteus are built on a flexible, open-software platform that includes integrations to a range of common instruments.” By supporting standardized data formats, this LIMS can perform in a wide range of lab environments.
Visualization could also enhance LIMS. In part, scientists want better ways to “see” what’s going on in entire processes. This desire triggered LabWare to explore visual workflows. “This lets a customer represent the entire work process on screen,” says Neri. “So, for example, LIMS can offer a person an interactive view of their lab’s actual floor plan depicting each freezer and the boxes or plates that are stored in it.” A dashboard of information could signal green: an instrument with room for more work, and red: machines out of calibration. “An dynamic image provides lots of latitude for automating workflow,” he says.
As vendors and scientists work towards standardized data formats and improving the tracking of research and results worldwide, LIMS can pull increasingly more complex sources of information into managing knowledge. As that happens, LIMS will grow from one component in a scientific process to a solid foundation that contributes to every step behind creating and understanding information.
Mike May is a writer and frequent contributor to Bio-IT World.
Subscribe to Bio-IT World magazine.