The completed sequencing of the human genome was a tremendous achievement, but only 30 months later, the long term knock on to the drug discovery industry is becoming apparent.  The prediction is that by 2005 just about half of all new products will come out of biotech, I presume that is why there are over 4,000 biotech-focused companies (approximately 10% of which are public) out there.  Throughout these organisations computer based research studies are increasing and notably, they are developing independent of the laboratory.  Well-integrated Information systems (IS) may one day develop into the primary R&D environment with analytical laboratories reduced to a secondary role of investigative analysis and manufacturing.  Consequently, I find myself wondering:

“How long before computer based research overhauls traditional biological research and development?”

Before I cause too much anxiety in the wet laboratory based scientific community, let me elucidate what I mean by computer-based research.  My view is that it includes but is not restricted to what many of you may already be thinking, namely the growing field of bioinformatics.  I would however broaden the scope to allow us all to become budding Informaticists, making use of technology, software and databases.  Scientists should be able to data mine not only external data sources but also their own unstructured data and pseudo-information in the quest for new discoveries.  We only have to think about recent trends of therapeutic switching with well-known products being recycled for secondary or even tertiary uses to appreciate how valuable informatics can be.  Two of the best-known examples being:

• Viagra (Sildenafil) - previously used to treat angina

• AZT (zidovudine) – Studied as an anti-cancer drug in the 1960’s

IS already play a significant role in most organisations and our industry is no different, in fact electronic systems are arguably more import in our situation than many others.  The volume of data that we all generate every day is exploding at an alarming rate; bioinformatics data requirements alone can bring some companies to a halt.  Being the area earmarked for expansion, it is, depending who you talk to either at hub of genetic research or the cause of all delays.  Of course, it is not just about data handling.  Everyone within the development process has a need to be involved and understand the decision making process.  Whether the requirement is for timely information to senior management, or the sharing of analytical and computerised data, the use of systems that facilitate the move towards computerised research are desirable.  In addition, validation and regulatory will continue to have a greater reliance on informatics and knowledge management, and satisfying their needs for the electronic IND (Investigational new drug application) submission with part 11 compliant pre-clinical test data is essential.

Industrialising the drug development process

The needs for adequate resources and to communicate effectively both internally and externally are critical success factors.  Why then when as this dependence on computers grows do so many new bio-medical companies fail to implement visionary IS strategies?  Surely, everyone is after some form of competitive advantage.  An IS strategy to complement the business plan and aimed at sustaining a maintainable competitive advantage by planning for the continued increase in computer research will certainly help achieve the goal.

In the business of drug discovery, I have observed a shift in information requirements as candidates progress down the development pipeline.  At the outset, the primary requirement is for the management of analytical data supported by peripheral information, but this gradually changes to one where the management and sharing of information takes precedence and is supported by analytical data.  One approach to smooth the progress of this shift in requirements is to “Industrialise the Drug Development Process.” 

As the company generate leads, then regardless of the source, scientists enter details into a database, tracking the analytical results and monitoring the different criteria that must be satisfied in order to progress the candidate.  This industry production-line parallel can provide a plethora of data showing not only development progress, but also statistics on where bottlenecks may exist and which of the many biological and computer sources are proving to be the most successful.  Taken even further and incorporating a document management system can provide management and scientists alike with a single dashboard interface to navigate the resulting corporate data. 

Many scientists argue that the effort required to maintain such systems is too great and that it distracts from their practical pursuits.  They should remember that ultimately it is the company’s goals that they must help achieve and should reflect on how many drugs or at least therapeutic uses are discovered and effective by chance.  Regardless of whether data is associated with early research or from later phase I or phase II pre-clinical trials concerned with safety or effectiveness, it can prove invaluable in later analysis.

Partnering with IT

If we place more emphasis on information systems, then we should keep in mind that simply having realms of electronic data is of minimal value and it matters little that there is a plethora of software and database tools out there to process the data.  To make efficient use of the nebulous data and in-house information, organisations need to collaborate with their IT departments and use the available technology to deliver effective systems.  Unfortunately, my familiarity of working with organisations of differing sizes, industry and culture, allows me to cite a recurring issue and that is that partnering with IT is always a problem.  Historically there is often a gap between IS/IT groups and the laboratory and even distrust.  Laboratories need to address this, because fragmentation of IT systems does not easily support a unified knowledge base.

“You may have noticed that the less I know about a subject the more confidence I have, and the more new light I throw on it.” -- Mark Twain

We can often attribute the “walls” that exist between the laboratories the business and IT to a lack of understanding of the industry, perhaps the regulatory constraints.  However, in my opinion a core contributor is the volume of data that we generate and the fact that we do not just need to store data, but rather push it between one another and share.  The lack of communication in conveying these needs is fundamental to why there is such difficulty.

The IS strategy should be in response to the business direction, but all too often the management team struggle to understand how technology can help and IT personnel fall short of appreciating what the managers and therefore ultimately the business wants.  Often an external contribution can help.  A business–IT hybrid who understands the business value of IT, does not see drug discovery and IT as separate entities and has hands-on experience in pursuing business objectives with information systems can bring both sides together.

Knowledge Management

While every organisation should be wary of too much knowledge being held in one person’s head, the issue here is much bigger than “corporate memory” and clearly more than using networked drives and digitising laboratory notebooks.  It is “knowledge management”; the recognition, storing and protection of all the information needed for the successful operation of the business.  Most start-up companies are already using knowledge management, people talk in coffee rooms and exchange ideas.  This informal exchange is vital but as organisations grow, this system like many others will prove not scalable and there is an increasing need for more formal systems.

Successful knowledge management will then rely on many things, depending on the size and exact nature of the business.  It is likely to utilise “search and retrieval” software against enterprise data systems and should always be part of an IS strategy. 

Common enterprise systems include:

·         LIMS – laboratory information management systems

·         DMS – document management systems

·         RDAS – raw data archiving systems

·         ERP – enterprise resource planning systems

You may already have some or all of the above, perhaps even multiple occurrences.  I wonder which of the existing systems were purchased as part of a clear strategy and more importantly how many you believe are fully contributing to the bigger picture.

The struggle then, must be for superior knowledge management with an optimised scenario where the organisation have pooled their resources having implemented a mixture of manual and electronic systems that allow people to share information, extract knowledge and stimulate thought. 

The way forward

Where do you start?  That is easy.  Firstly, define the factors that are critical to the success of your business and then consider the consequences of costly mistakes that are often repeated and the cost of opportunities that are often missed.

I wish you all the best of luck.  We are only talking about the very beginning of the development pipeline here, but maximising added-value to this early stage makes real sense when considering the hundreds of millions of dollars that it can take to get a product to market.

About the author