Thought on the Electronic Lab Notebook

This topic is becoming more interesting to me every day as we see the technologies and the processes evolve. A simple definition of an Electronic Lab Notebook (ELN) is becoming almost impossible to agree on as can be seen in a recent discussion group I am a member of over on LinkedIn that points to the LimsWiki definition.

An electronic laboratory notebook (also known as electronic lab notebook or ELN) is a software program or package designed to replace more traditional paper laboratory notebooks. Laboratory notebooks in general are used by scientists and technicians to document, store, retrieve, and share fully electronic laboratory records in ways that meet all legal, regulatory, technical and scientific requirements.^[1] A laboratory notebook is often maintained to be a legal document and may be used in a court of law as evidence. Similar to an inventor's notebook, the lab notebook is also often referred to in patent prosecution and intellectual property litigation. Modern electronic lab notebooks have the advantage of being easier to search upon, support collaboration amongst many users, and can be made more secure than their paper counterparts.

A mouthful and it scrapes the surface of what may be meant when speaking of an investment in an ELN. I gave a talk at an informatics conference in London a couple of years back and showed a chart with overlapping functionality intersecting the role of traditional LIMS systems and the growing ELN segment. That chart has moved forward to the point where now you find markets where they are almost one and the same.

This brings up a couple of interesting discussion points

• Do we need both, or should we blend the systems to make it one experience?
• What is the value of a lab notebook? (GxP / Exploratory / Other – context is a big part of this but I am thinking beyond those aspects to the scientific thought process)
• Does the act of capturing free thought in a (paper?) notebook encourage a different type of thinking than the act of capturing data in a structured system that allows encourages free text or unstructured input as a part of the structure?

Clearly there is room for a discourse on all these topics, and many more that dive deep into the specifics, but as we look at the tools, I can’t help but come back to the question of what are we really trying to do? The technology is not the barrier to capturing data, or even in most cases, information. But are we making the technology a barrier to the process of creative thinking?

LIMS are by design typically built to capture structured information and results sets, manage samples and lifecycles and other common analytical workflow. Paper lab notebooks have for centuries been used to record everything from wild ideas that need to be formed into hypothesis and examined before even approached as a test, to being used as evidence of test execution or development. Now, in the litigious world of drug research and development, they hold a very high burden depending on where in the lifecycle they sit, acting in some cases as patent defense platforms!

I wonder how much we have moved away from creative thinking and free capture of those ideas in our pursuit of the right toolset to capture our data. I know a common theme I have heard from my scientific associates over the years is that there is less and less time for thinking, and more and more effort devoted to execution. Looking at the innovation cycle and the pharma pipeline as a whole, I wonder sometimes how much of the hollow sound in that pipe is regulatory pressure, or external pressure, and how much is a shifting in innovation and creative thinking.

Social Mention

A while ago I ran across a tool called Social Mention. The idea is pretty sound. You can see the description from the site below. The applicability extends across industry and makes a pretty interesting marketing research platform in it's early stage.

Social Mention is a social media search and analysis platform that aggregates user generated content from across the universe into a single stream of information.

It allows you to easily track and measure what people are saying about you, your company, a new product, or any topic across the web's social media landscape in real-time. Social Mention monitors 100+ social media properties directly including: Twitter, Facebook, FriendFeed, YouTube, Digg, Google etc.

Chemistry Add-in for Word

So as promised, here is the blurb on the Chemistry add in for Word. Very cool idea and I hope they include it in PowerPoint as well since so much of the work ends up there for presentation.

The Chemistry Add-in for Word makes it easier for students, chemists, and researchers to insert and modify chemical information, such as labels, formulas and 2-D depictions, from within Microsoft Office Word. In addition to authoring functionality, Chem4Word enables user denotation of inline “chemical zones,” the rendering of high-quality and print-ready visual depictions of chemical structures, and the ability to store and expose semantic-rich chemical information in a semantically rich manner.

You can head over to the MSR site and grab it here.

BioCoder - code for Biologists

While trolling through some Microsoft research projects recently I ran across two of particular interest in my current area - BioCoder, described below, and Chemistry for Word which I will cover in a later post. The BioCode concept applies programming constructs to the protocol definition process. The idea is interesting and worth exploring. The blurb below from the site summarizes it nicely and you can read more by downloading the package here.

BioCoder is a high-level programming language that enables standardization and automation of biology protocols. Our vision is to change the way that experimental methods are communicated: rather than publishing a written account of the protocols used, researchers will simply publish the code. The code can be automatically converted to human-readable steps for manual execution in the laboratory. When written as a computer program, biology protocols can be parameterized to facilitate reuse in different contexts. They can also be mapped automatically to the setup of a given laboratory, taking into account the equipment and reagents that are available.

ELN Talk this summer / fall

I'm scheduled to give a talk in Amsterdam billed as "The leading conference for ELNs and Data Management strategies" at the end of the summer. The details are online at the conference website.

My talk is: Keynote Presentation: The Role Of ELNs: A Chapter In A Much Larger Story

• How do you strategically select an ELN to fit in with your company requirements?
• Outlining the importance of creating the enterprise strategy to create a seamless informatics platform
• Delving deeper into the Wyeth/Pfizer GLP ELN deployment: why was it a success and what lessons were leaned along the way

I enjoy these events as I almost never leave without great contacts and networking. Any thoughts to contribute are welcome, so feel free to get in touch with me.

iPads and ELN's

It seems the iPad is everywhere lately, even popping up on a LinkedIn discussion on Lab Notebooks.

My opinion on this topic is essentially the same as the tablet-pc form factor. The tech is not there yet to get any real advantage in most cases, though there may be a few isolated edge cases. What I think we are waiting for is the next disruptive way of recording lab observations and tying the information flow together in a more seamless fashion. I do not see anything the iPad will make easier yet, except the type of task already relegated to the iPaq, or similar form factor which is essentially basic, situation specific, observational recordings. Now - don't get me wrong, I think it's very cool and would like to have one dropped in my lap to play with, but from a business value in the labs perspective, I think we are waiting on the killer app. (Let's just hope it's not Flash based)

Corporate cost cutting

Today was the big day - sort of. We heard the first of many site announcements though these will be the biggest I am sure. Turns out my client base is all on it's way out, making my role as a Business Partner a little redundant. It will be interesting to see how long things take to complete and what the transition will look like. My integration work will shift to more of a transition role I imagine and my other work will take on new importance as that will be the primary focus once the transition work is complete. Good luck all my friends in preclinical / pre-poc!

Linked Data

The concept of linked data is not new, but is finally hitting the tipping point for implementation where thought leaders, technology, and business drivers are coming together to bring it from theory to reality. Tim Berners-Lee gave a great talk on this topic at a TED conference. Check it out here.

I was moderating a talk at the recent AAPS conference in Seattle and one of my panelists was Randy Julian, from Indigo Biosystems. He focused his talk on the value of linked data in the pharma research space and how it can revolutionize the collaboration process. I very much enjoyed the talk and our subsequent, but brief conversation. We agreed to talk more and I had to bolt as I had meetings with Microsoft and had to catch a plane for Europe to speak at another conference. Things continued to come together with the meeting at Microsoft. While the discussion was brief, it centered around the Amalga platform and the underlying technology. The aim is very similar, in that it looks at data at the atomic level with appropriate meta data for context and just in time reassembly as needed, inferring relationships automatically. The implications for predictive and retrospective automated analysis are huge.

Later that week, I spoke at a conference in Budapest and caught a talk with Nico Adams, from Cambridge University in the UK. He talked about the Semantic Web and how it applies to the research area. We later had dinner with the larger group and talked a bit more. It is encouraging to me to see the synergy across the world, and it points to the critical mass that is building around this topic.

Budapest

The conference went well and I got a solid response on my talk. The morning talks all pointed toward foundational needs for deploying these solutions and set up the discussion quite well.

Today, I caught a talk from Nico Adams, of Cambridge University, regarding the Semantic Web. It was a good talk and what really got me was that this was a follow on to a great talk by Randy Julian, CEO of Indigo BioSystems, at the AAPS earlier in the week on a very similar note, and also followed on the meeting I had with Jim Karkanias at Microsoft regarding their information strategy and the thinking behind the Amalga platform. While they are all on different paths, they are headed in similar directions and the thinking trend is toward data management at the atomic level. I am looking forward to seeing where this goes.

Sunday Talk at AAPS

The talks went well Sunday for my session with Joel Usansky from Thermo, Jeff Tishler from IDBS and Randy Julian from Indigo BioSystems. The first two talks were predictably product focused, with some tie in to Ligand Binding Assays and general Bioanalytical work flow. When Randy stepped up, he led off with a slide contrasting vacuum tubes and transistors, highlighting the need for something new, as opposed to optimizing the current systems. He then went on to talk about linked data as opposed to the standard static schema architecture of our database systems. He lost a good portion of the audience as evidenced by some of the questions, but the talk was spot on and tied together several ideas talked about earlier.

I am looking forward to following up with Randy when I return from Europe to compare notes.

AAPS - Update from Seattle

The speakers dinner last night was enjoyable and informative. The challenges coming out of the ligand binding labs are the same as those faced by the rest of my clients and center around key information exchange challenges.

As a deliverable set from yesterday, the group discussed several key focus areas, several of which resonated with me.

Data exchange was a hot topic with a desire for standardized information exchange so systems could talk to each other with minimal work. There is already work in the industry on this topic and I pointed to several of the current and emerging transport standards.

Instrument and software validation and the easing of the process was a highly debated topic as it represents so much of an impact for the GLP labs. The options and discussion ranged from validation accelerator packs to FDA recognized validation standards that allow a vendor to certify, reducing the customer level validation to the 20% or less of customization or configuration.

Also discussed was the desire to have a standardized "plug and play" infrastructure for instrument hardware. This spurred a number of discussions around the implications, not the least of which is the need to implement bidirectional communication in the hardware and software layers of the instruments to make this possible. The lack of common standards to build to and a governing body to work with makes this all the more challenging. My suggestion to the group was to not take on this as an outcome, but rather take on a paper with requirements and standards recommendations, that can then be used as a foundation for discussion with the key players. We need to get on the same page with respect to what it means to have integration at the hardware and driver level before we can get to the level of discussing implementation in detail.

All in all, it was good discussion and I am looking forward to my session later today, discussing the ELN and LIMS roles in these and other areas.

Traveling...

Philly to Seattle, Speak at a conference, meet with people. Seattle to Frankfurt, Frankfurt to Budapest. Speak at a conference, meet with people and see the city. Budapest to Frankfurt and Frankfurt to Philly - home again. A lot of time in the air is in my future this month!

Ontology Lookup Service

Who knew - Sourceforge supports an Ontology lookup service! Check it out here: http://www.ebi.ac.uk/ontology-lookup/

I have to dig a little more to see if there is a publicly accessible API we can hit for this - sort of like encoding a-la MEDRA.

The ELN One Vendor Approach

Having heard the story too many times, I had to write a note about the vendor / application selection process for ELN. There seems to be a prevailing thought that a monolithic, one vendor approach is the way to go, regardless of business application detail. To understand the pros and cons, let’s briefly examine the foundation of this argument.

One vendor is easier to negotiate with, allowing for greater economies of scale and commonalities in infrastructure and support. With one application, support analysts can be deployed to cover an entire business unit and with the greatly increased scope, realize the benefits of scale. In summary, a one vendor approach is an IT persons dream, right?

Oh - wait, it's not all about IT? You mean without the "Business" there would be no need for IT? Ahhhhh! Well then, that changes things a bit.

Now it's time to start thinking about fit to purpose / fit for use. There is currently no singe ELN vendor that covers all the spaces, though many make the claim. How similar is a process chemistry requirements set to a small molecule biologics work flow? How about GLP vs. Non-GLP? Discovery work flow and the wild west process management has a bit of a different set of needs than a Regulated Bioanalytical group.

In general, ELN purchases must be carefully thought out to match the needs of the users and deployed as a targeted asset, not a generalized commodity application. While it is clear you want to avoid the "one of everything" mentality we sometimes get in large pharma, we need to accept that this space is still a multi vendor play for the foreseeable future.

Frustration with vendors making claims about “knowledge management”

We have a system that can uncover all types of hidden relationship and turn your data into knowledge! It is starting to sound like a carnival hucksters line to me. Once the conversation gets underway, I start to ask a few clarifying questions and things begin to become both clearer and muddier at the same time.

It becomes clear that the claims are over hyped, and at the same time, separating the actual out of the box capability from the hype or customization becomes muddier.

Too often, these systems rely on an underlying established dictionary, specific ontology, and / or custom meta data repository. This often has to be uncovered by direct questioning, and it is critical to clarify what comes with the product vs what is custom built for my business area. How does your established dictionary and overlying ontology map to my process? What about linguistics, allowing the tool to anticipate what I need based on my query and to learn from my results? I am ok with a learning curve if I see where we are going.

Over hyping the capabilities with pre-set cases only sets up the pilot / deployment for disappointment. When users see a case that looks like magic, they expect magic for their systems. Bottom line – sell what you really have, wait to market your “vision” until you have something real. It’s ok to say, here is where we really are, out of the gate. Do not waste my, or others time with misleading claims.

Ok, i'm off the soapbox... for now.

Pleased with the results

We had a good meeting with Pfizer today. The team was good and seemed genuinely interested in fully understanding the landscape at Wyeth. The repeatedly stated concern was that they capture sufficient detail to not leave any business users in the lurch when things went down. All in all, despite the building and meeting room shuffle, it was a productive visit. I am sure there will be many more, across many more areas, to get things fully sorted out but the process seems to be on track so far.

The announcements on organization today were also an interesting step – it looks like the executive level will get some Wyeth participation in the new organization. While the roles named were all key business folks, I think they reflect a desire to maintain the talent that lead to the acquisition in the first place. Perhaps that’s my eternal optimism surfacing again, but either way, I am feeling upbeat about the process.

The looming specter of the 20,000 projected layoffs is certainly on most everyone's mind as we go into this process, but the best way through is to focus on the solutions and while planning for the worst, look for the best. The drain of worry can become a self fulfilling prophecy if you allow it. - no extra charge for that bit of advice ;)

Headed to NY

I'm on the Accela Express, headed to NY to meet with Pfizer around our upcoming integration activities. This is a continuing dialog, involving a significant amount of information exchange and effort by a large team of people. My hope is that the effort being put into this integration will be reflected in the results, with a smooth integration of pipelines and people.

The history of these mega mergers tells a different story, with pipelines and productivity taking a hit, accompanied by a serious "Brain Drain". The stated effort is to make this one different. I am hopeful - though I am the eternal optimist while planning for the worst.

Ontology based data & semantic relationships

Working through a data architecture and strategy for clients recently, I had to compile some information regarding development of the landscape. This information is reflected below.

The primary components to reference are detailed in the included diagram. In this case, the ontology is clearly a piece of the stack, but not the “data” or the single UI.

The ontology is best thought of as a view of an established information set that uses concepts to define relationship. The data is then mapped onto ontology to provide a specific high value view of the data, aiding in the generation of information and knowledge. Clearly there is an assumption that as a precursor to the use of the selected ontology, significant work has gone into the process of cleaning the target data through direct manipulation or a meta data based transformation layer to manage synonym matching, etc. across sources.

Another definition, more succinct is, “Ontologies are computable conceptualisations of a knowledge domain” as defined by Nico Adams, emphasizing the transformation from data to information that can then be further acted upon.

The OWL Web Ontology Language is designed for use by applications that need to process the content of information instead of just presenting information to humans. OWL facilitates greater machine interpretability of Web content than that supported by XML, RDF, and RDF Schema (RDF-S) by providing additional vocabulary along with a formal semantics. OWL has three increasingly-expressive sublanguages: OWL Lite, OWL DL, and OWL Full. Source: http://www.w3.org/TR/owl-features/

The semantic layer-cake.: (Copyright © 2008 World Wide Web Consortium, (Massachusetts Institute of Technology, European Research Consortium for Informatics and Mathematics, Keio University). All Rights Reserved. http://www.w3.org/Consortium/Legal

Useful links :

1. W3 OWL Reference a) http://www.w3.org/TR/owl-features/
2. Good article on Ontology implementation a) http://www.sciencedirect.com/science?_ob=ArticleURL&_udi=B6T64-4GMB0F0-F&_user=358874&_rdoc=1&_fmt=&_orig=search&_sort=d&view=c&_acct=C000017638&_version=1&_urlVersion=0&_userid=358874&md5=4b5c6decc4b2f01fad8183b00aec72b1
3. 3) DDI - An Ontology for Drug Discovery Investigations a) http://users.aber.ac.uk/ddq/ddi/wiki/index.php?title=Main_Page
4. Open BioMedical Ontologies a) http://www.obofoundry.org/crit.shtml
5. OWL ontology browser a) http://pellet.owldl.com/ontology-browser/
6. OBO Ontology Download Matrix a) http://www.berkeleybop.org/ontologies/
7. Semantic Chemistry a) http://www.semanticuniverse.com/articles-semantic-chemistry.html

Reference:Stephen P. Gardner, Ontologies and semantic data integration, Drug Discovery Today, Volume 10, Issue 14, 15 July 2005, Pages 1001-1007, ISSN 1359-6446, DOI: 10.1016/S1359-6446(05)03504-X.(http://www.sciencedirect.com/science/article/B6T64-4GMB0F0-F/2/f296f93f44347cfd561668ede72ac5f9)

Nico Adams, Semantic Chemistry http://www.semanticuniverse.com/articles-semantic-chemistry.html

More to come...