Q Where will bioinformatics be in five years? Ten years?
A The next five years should see great progress in supplying databases and search tools with automated annotations that are driven by links into scientific papers. The point is to gain knowledge from data and this is not easy. The first step is to make sure that as much of the relevant data relating to a particular inquiry is assembled in a single location. The second step is to allow for very rapid inquiries driven by a user who is provided a means to assemble a picture, literally and figuratively, of the underlying biology. Then the final step is to have an increasingly large fraction of these enquiries take place by automated means, assembling the pictures for the user prior to their manual intervention or specific request. All this may take 20 years to really work as people are envisioning it today.
Q What are the biggest challenges bioinformatics must overcome?
A The largest challenge will come from bridging fields and disciplines in a unified search environment that avoids being tedious. In a given set of queries one would like to combine both clinical outcomes data, economic data, expression data, genetic variability data, and proteomics data all in a given session. Combining diverse sets of information and displaying it in a meaningful fashion without several minutes of search time is a key threshold for maintaining mental momentum in the pursuit of truly innovative enquiries. The pay-off from this is to enable an entire population of diverse scientists to access each other’s domains and have answered questions that they can barely articulate.
Q What hardware do you use?
A We use a variety of platforms and are continuously reviewing our requirements. We will soon be investing in hardware and informatics infrastructure. Much of the investments will be for customization of existing hardware and software solutions.
Q What bioinformatics software do you use? Do you use in-house developed or third-party software?
A We are currently reviewing both bioinformatics as well as clinical trials database tools. We assemble and customize most of our software and only create de novo programs when they are core strategic applications relating to SNPs.
Q How do you integrate your data?
A We currently do this manually. Once the basic correlations have been found using standard database techniques, then much of the work can be done with simple software tools. The next level of integration is to access the large clinical data repositories that we are creating and gaining access to. We are not sure that the integration with state-of-the-art bioinformatics and clinical data management has to be seamless for a great deal of progress to be made.
Q Whose DNA chips do you use?
A We use Affymetrix chips as well as our own microchips that we have produced internally.
Q How large is your bioinformatics staff? Is the company hiring additional bioinformatics staff?
A We have approximately five to 10 people actively engaged in bioinformatics/clinical informatics infrastructure and double that engaged in daily interactions with the data systems and data. We will be doubling this resource over the next several months and probably re-doubling next year.
Q How is your bioinformatics unit organized within the framework of the company? Is it part of the IT department, the biological research department, or some other department?
A Currently, we have Web-based commerce, corporate-oriented, bioinformatics, and pharma clinical trials informatics all as separate organizations. We are now at the early stages of considering reorganization so that these groups are more closely aligned, and some form of reorganization is likely.