Where will bioinformatics be in five years? Ten years?
Bioinformatics is currently a computational science, dominated by programs and algorithms that are run repeatedly and identically by different labs. This setup has been necessitated because new data is entering the system, and must be fit into the whole. The next phase will be the ability to look up any human gene from a database.
What are the biggest challenges bioinformatics must overcome?
The amount of data that will need to be held, the mechanisms for accessing the data, and the checks and balances for changing and updating the data.
What hardware do you use?
We are primarily a Microsoft Windows NT-based company for all of our desktops and corporate servers. We also stick with the Intel platform for most Unix applications, using Linux. But we also have a few Sun Microsystems servers for our laboratory information management system.
What bioinformatics software do you use? Do you use in-house developed or third party software?
Our bioinformatics department is composed of tool makers, not tool users. The tool users are the scientists in the lab, and they use a combination of tools that we develop or that they buy from outside. We have experimented with most of the common packages. We had GCG for a few years, but nobody used it. We now use various best-of-breed PC products and various websites like that at the US National Center for Biotechnology Information. We are interested in using an application service provider with a Web interface, but have not settled on one yet.
How do you integrate your data?
We save data in Oracle databases using our own schema. Integration becomes an issue of moving data in and out of Oracle. For that, we use scripts written in Perl or Tcl/Tk. Our data layer standardization is based on the Open Database Connectivity standard to keep us the most flexible. We are using SAS as a statistical layer on top of our database, and we are very impressed with the company’s ideas for data warehouse and reporting functions.
Whose DNA chips do you use?
We have our own MassArray system for genotyping, which has special chips designed for mass spectrometer analysis.
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?
Our bioinformatics department is closely aligned (both physically and organizationally) with the information technology department. Our customers are the molecular biology and genomics departments at Sequenom, who have their own ex-perienced users of bioinformatics software and tools.
What non-existing technology do you most wish you had?
The answer here is easy, but may surprise you. The biggest missing capability we have found is in database reporting tools. We thought that LIMS vendors would have solved that problem, but none of them have--they use third-party tools. Some even said they had tried to write a reporting tool and failed. We have also tried and so far failed to attain the requisite balance of flexibility and power--it is not easy to allow the user to specify what they are looking for without exposing them to the underlying schema.