MOSS in Action: The Scientist’s Workbench
I recently attended the "Microsoft Health and Life Sciences Industry Council - The Scientist Workbench" in Malvern, PA (http://blogs.msdn.com/allandcp/archive/2007/12/18/microsoft-health-and-life-sciences-industry-council-the-scientist-workbench.aspx). The Scientist Workbench is another great use of SharePoint, applying the technology to solve real-life problems and save money. The business case for using the Scientist Workbench (a framework that sits on top of SharePoint to help drug researchers visualize, collaborate on, and search data better) goes something like this. The following conditions are often true:
- Most drug companies tend to be organized across therapeutic areas.
- The drug discovery process tends to be more of a silo than a collaborative, information-sharing process, between therapeutic areas and even between stages of development in a single therapeutic area
- An ELN (Electronic Lab Notebook) starts to address this, but there is much more benefit to be had
- Pipeline books detailing which compounds are at what stage tend to take a while to collate, up to two months (forcing executives to make decisions with stale data)
- The longer a research project lasts before trimming it (if the drug is not effective in treatment), the more money is spent in vain.
- The lab process, be it using real experiments or "in silico" experiments is often rife with opportunity to automate data acquisition, transfer, scrubbing, and reporting, reducing the time and increasing the efficiency of expensive research PhD's
- Co-development between drug companies presents unique security challenges of sharing all the data that is needed for a project with the right people on both sides, and hiding all other data
- What if your next blockbuster drug is sitting in your discard pile right now? (see the famous story about the hypertension drug that grew up into something else entirely http://pubs.acs.org/hotartcl/mdd/98/novdec/viagra.html)
The Scientist's Workbench is a framework for addressing these problems. Note that it is not a shrink-wrapped product, it is a tool for companies, integrators, and ISV's to develop solutions. Since every company tends to have slightly different processes or tools, it is nearly impossible to deliver a full-blown product that meets every permutation of needs across the space. The framework instead allows an organization to quickly develop a solution using SharePoint. Some examples of how the power of the workbench can be harnessed to address the issues above:
- 1) An OBA-style solution can quickly be developed to provide an ELN solution that uses the client-side tools scientists are used to working with (Word, Excel, etc) and SharePoint as a storage mechanism on the server. Metadata can be captured along with real data. Data feeds from lab equipment (even better if it emits XML) can be integrated as well. Add an extra 10-15% of efficiency just in the lab process using SharePoint and Excel Services!
- 2) Building on the first step, this data can be made searchable (with controllable security) across the stages of development, and/or across therapeutic areas. Additionally, this data can be shared outside the firewall with a co-development company if needed, with item-level security and access controls.
- 3) The time to collate a pipeline book can shrink to days from months, allowing near-real-time decision support to executives
- 4) "Fail early" and discover sooner if the efficacy of a drug is sub-par, saving research money.
- 5) Allow you to re-examine compounds down the road, and understand the complete data and context of your work months or years later.
- 6) Make your compliance efforts (21 CFR Part 11, Sarbanes-Oxley, HIPAA, etc) easier using MOSS with A/D, Records Center, and Information Rights Management
The Scientist's workbench contains a whitepaper and presentation, a sample implementation, and a MOSS template. It is based on the Connected Life Sciences Framework, and is promoted by the BioIT Alliance (http://bioitalliance.org/). It will be released soon as Open Source, when it comes out I'll update here or in the comments with a URL for download.