1. The system is very old and doesn’t work well.

2. You don’t actually know what all your system is doing.

3. Your system is s data graveyard, a place where the data goes in to die and never to be able to be used for data sharing and knowledge management.

4. The system is a critical one and its technology is out-of-date.

5. Your needs have drastically changed since you purchased it.

Actually, I just copied these points from an older article of mine, but the article is a lot worder than just reading these points in this short form:

http://geometrick.com/n-l-how-to-know-when-upgrade.html

• The commercial products are more mature, although the marketplace is still very dynamic with merger and acquisition activity, and new players coming on to the scene.
• Electronic records can be used as evidence in court.
• ELNs really can be used for biology.
• The ‘Cloud’ is a remote server farm.
• Our vocabulary is being extended to include words such as Semantics and Ontologies.
• We need to share data and information across the firewall.
• Thin clients become increasingly important.
• ELNs can be as portable as paper notebooks.
• The need for standards is increasingly recognised.

But there a few things that remain unresolved:

• What is the difference between an ELN and a LIMS?
• What is a Laboratory Execution System?
• How secure is the Cloud?
• Will we ever solve the integration problem?

And of course, there are the permanent concerns about best deployment strategies, return on investment, and getting user buy-in.

Click the link for the programme for ELNs & Advanced Laboratory Solutions. The Download Centre contains a number of articles, presentations, podcasts and videos aligned to the ELN event.

The event programme is available from the link above, but here’s a few of personal observations and highlights from the presentations. Burkhard Schaefer gave an update on progress with the emerging AnIML data interchange standard, placing an emphasis on the cost savings that can be achieved through the elimination proprietary approaches. One specific design feature of the AnIML standard is the provision of a generic data container (AnIML Core) which offers the storage, not only of the analytical data, but also sample information and method information, including the instruments and software used. It is interesting to speculate to what extent this container could serve as the core of a new type of LIMS or ELN where the complete sample or experiment record exists entirely independently of the ‘system’ that created it. There’s a podcast of an interview I conducted with Burkhard Schaefer here.

Three presentations from Philip Laffin and Jerry Lanfear of Pfizer, Claire Wilson and Fionnuala McDyer of Almac Diagnostics and Erik Junius of Waters, each touched on process improvement in the laboratory. In each case, the application of a systematic process for breaking down workflows and identifying and eliminating bottlenecks led to time and cost benefits.

Robert Davey (The Genome Analysis Centre) and Mohammad Afshar (Ariana Pharmaceutics) both addressed issues about data volumes and data analysis. The exponential growth in the output per day of kilobases per sequencer (projected to reach 1 billion per day in the not too distant future),  is presenting the Genome Analysis Centre with mind-boggling volumes of data.  The presentation described how an infrastructure is being constructed  to be able to cope with these data volumes.

Managing clinical images based on the DICOM standard was a topic addressed by Teodor Wagner (F.Hoffmann – La Roche) and Omer Casher (GSK Clinical Imaging Centre). A common thread from both presentations was a concern about vocabularies and metadata, particularly when multiple data sources are providing images. Even if the image is in a standard format, its true value can only be realised if standard vocabularies and metadata can be applied.

Ulrik Nicolai de Lichtenberg (LEO Pharma) gave details of a project to integrate biological and chemistry data based on the Isentris (Accelrys) platform. Of particlaur interest was the adaption of Maslow’s hierarchy of human needs (a sociological construct) to a similar hierarchy of data management needs; the proposed tiers in the model (imagine the bullet points as a pyramid):

• Advanced visualization, analytics, modeling (across all data domains)
• Integrated search & browse across domains
• Information is structured and searchable
• Information can be found and shared
• Information should exist

Presentations from Damien Voisard (Merck Serono) and John Leonard (Astra Zeneca) both dealt with process improvement related to manufacturing. Merck Serono utilises a Process Data Management System that enables early fault detection that has been shown to significantly enhance process knowledge. The emphasis from John Leonard was rapid process transfer from R&D to manufacturing which has lead to a reduced number of manufacturing documents and a >50% time saving in the preparation of manufacturing documents, improved quality, transparency and availability of documents. Improvements in time to market were also the subject of Mike McGorry’s (PDMS, Johnson & Johnson) presentation, based on the International Society for Automation (ISA) S95 (integration of enterprise and control systems) and S88 (definition of production processes) standards.

The roles of LIMS and ELNs in complex business landscapes was the subject of a presentation by Mark Gonzalez (LabWare), touching on issues of integration, particularly in the situation where the LIMS and ELN are basically part of the same system.

Although I’m sure the organisers are disappointed with the relatively low number of attendees, the meeting offered a broad, high quality view of a number of background issues facing laboratory data and information management, focussing on the problems rather than the tools. The underlying theme of integration and (lack of) standards is often at the heart of this type of meeting, and this became a topic of discussion in a brief open forum session, giving attendees the opportunity to express their opinions, and frustrations, but only serving to emphasise the lack of any collaborative approach towards resolving these concerns.

Elisabeth Goodman, of RiverRhee Consulting, has already posted an excellent report on the Exchange, which you can read on her blog, Operational Excellence and Knowledge Management in an R&D Laboratory Environment. Aware of the fact that Elisabeth had to leave before the final two papers, all I can add to her comments is that the AnIML project (Analytical Information Markup Language) was the subject of a presentation by Burkhard Schaefer, aided and abetted by a case study of AnIML in action by Mike Ludlow of LGC Pharmaceutical Services.  See AnIML – An interview with Burkhard Schaefer for information about AnIML.  AnIML is an XML-based standard that is developed under the umbrella of ASTM with participation from a number of different stakeholders in the pharmaceutical and environmental industries, government agencies, instrument vendors and academia.

The final presentation, by Gareth Maxwell of Northern Ireland Water, addressed the use of LIMS in a highly regulated, but non-pharma environment. Gareth’s presentation was focused on the well-publicised problems experienced with the water supply in Northern Ireland last December, caused by some dramatic and rapid changes in weather conditions.  The scope and scale of the testing operations needed to support the delivery and maintenance of water supplies is extensive and subject to regulatory scrutiny from four different agencies.  The adoption of a number of technology-based improvements, all interacting with the LIMS, have helped bring about time savings and error reductions in a number of operations.

Just a word about my think tank session on ‘Overcoming the most common challenges facing laboratory integration’. This session was run as a discussion, which I kicked off by asking the question ‘If we were to start anew to build an integrated data/information solution for our laboratories, would it look anything like the solutions we have in place today”. The question drew some different responses, but a common thread that seemed to emerge was that the lack of standards was a key issue. In a short session, we were never going to cover much ground in the discussion, nor, indeed, get to hear all opinions, but the overall consensus seemed to point towards the need to separate content from the tools (applications) in order to facilitate the use and re-use of the content in a multiplicity of best-of-breed tools. Standard data interchange formats would therefore be the catalyst, requiring each tool to have the capability to work with those formats. The concern that standards stifle innovation in technology development is a risk that was expressed, and acknowledged, but in the main, the content is not the subject of technology innovation. So, my opening question failed to get a clear answer, but (a) I never expected it to, and (b) I’m not sure there is an answer, yet…..

Overall, I found this another successful event in the series; well attended, high quality content, and high quality interactions. In addition, the location, within walking distance of a number of iconic sites of modern history, not only gave delegates the opportunity to get some (very!) fresh air, but also presented an opening gambit to start conversations. The programme itself spanned a number of topical subjects, but if anything, it is laboratory integration that seems to be rising to the top of the agenda and the area of most concern.

Matthew makes the point that other industry sectors have seen the establishment of representative associations to establish and promote the adoption of standards.  It would be unfair to say that this hasn’t happened in the laboratory world, but sadly, few of the worthy attempts to address standardization have gained sufficient inertia to have had much impact.  Interestingly, most people involved in laboratory informatics and information management bemoan the fact that we do not have appropriate standards.  The only counter argument I’ve come across seems to suggest that standardization can inhibit innovation.  Well, that may be the case in certain sectors, but the innovation that we are looking for in the laboratory is in the science, and the capability of the tools that we need to gather, process and interpret data, not in the format of the data.  Like a lot of people in Life Sciences, Matthew is looking towards the Pistoia Alliance to provide the initiative to resolve this problem.

You can listen to the podcast here.

AnIML is an emerging and open data standard, maintained by the ASTM E13.15 sub-committeee on analytical data. The XML-based standard is independent of measurement technique, and supports the sharing and storing of analytical data. AnIML consists of a core data container that includes sample, method, instrument and software information plus measurement results. Extensions to this generic container facilitate the application of the standard to any established technique (chromatography, spectroscopy, etc.) as well as custom or one-off experiments.

Now, data standards in themselves aren’t really that exciting; it is the opportunities that they create that should get us really interested. So what does AnIML offer? Data interchange is an obvious benefit; the ability to transfer data, analyse it, view it, manipulate it, re-process it, independently of proprietary software and vendor lock-in represents a an unprecedented degree of freedom. But the less obvious benefit is the ability to archive data in a human-readable format over the long term. Moving to open, rather than proprietary formats has a multitude of other benefits as well, such as cross-technique data mining, data publication, LIMS and ELN integration, and other forms of result delivery.

I was also able to see a demonstration of the Seahorse Scientific Workbench from BBSN. At this point it is important to stress that theintegratedlab.com is an independent website that does not support the promotion of products or services, and my comments on the Seahorse Scientific Workbench represent a strategic view of a different product concept.

The Seahorse Scientific Workbench is an open and extensible ‘platform’, or software suite, that provides a number of data management and manipulation facilities for analytical data based on the AnIML standard. It offers a workflow driven approach that can complement, or offer an alternative to, other laboratory applications.

Without getting into the detail of the workbench, perhaps the two things that struck me most were:

1. An open-based approach starts to offer realistic alternatives to established proprietary systems to meet the demand for data interchange, sharing, collaboration and long-term data preservation.
2. The ‘workbench’ concept creates an opportunity to adopt a more modular approach to laboratory computing. For example, at the core of an ELN is an authoring tool that provides the capability to create a compound document. Most of us already have tools with this capability on our desktops. But almost every ELN installation will require integration with other laboratory tools, hence the range of highly functional, commercial products in the market place, that allow integration with chemistry, biology and analytical systems. If a modular, workflow-based ‘workbench’ facilitates the integration of other laboratory systems, then it offers an alternative approach to the traditional ‘application-centric’ approach that most laboratories adopt. This may not, of course, suit some organisations, but can offer a viable alternative to a typical data acquisition/LIMS/ELN configuration.

I will be conducting a podcast interview with Burkhard Schaefer ahead of two upcoming European conferences, at which he will be talking about AnIML. So those of you who were unable to take advantage of a trip to the sunshine and warm temperatures of Palm Springs can catch up at the SmartLabs Exchange in Berlin, Feb 28 – Mar 2, and Laboratory Information Management in Amsterdam, 23-25 May.