AT A GLANCE: Holds a bachelors degree in biochemistry from Middlebury College and a master of science degree in biotechnology from Tufts University. Interests outside the lab include playing hockey, softball, soccer, tennis, and golf.
QWhat role do microarrays play in research at Whitehead?
AA variety of projects by Whitehead scientists involve microarrays. The classic application is expression profiling. By looking at how cells respond at the transcriptional level to various stimuli we can learn a lot about how genes are co-regulated and how different pathways work together.
QHow is the microarray facility set up within Whitehead?
AIf someone has a project related to microarrays, they do the biology, including growing the cells, harvesting RNA and labeling targets. We make the arrays, and perform target hybridization and data extraction. We also help with the downstream data analysis.
QWhat types of microarrays do you use and in what combination?
AMost of the arrays we use in this facility are spotted here. The probe DNA we use is either PCR-generated cDNA or synthetic oligos. In addition, the group also uses commercial arrays from Affymetrix and Corning. This facility was started as a collaborative effort between the Institute and Corning to develop microarray technology. As part of that agreement we do beta testing of their arrays and provide them with a valuable outside user perspective. We also provide them with input for potentially interesting content for arrays, and work to develop novel technologies or applications that we can develop under the Whitehead Institute/Corning consortium.
QWhat kind of arrayer do you use to make your own arrays?
AWe have a Cartesian robotic arrayer that uses quill pins.
QWhat is the biggest challenge you face in working with microarrays?
AConsistency and background noise. One of the fundamental issues at the data analysis phase is how to deal with [these issues] in a strategic way.
QHow do you tackle these challenges?
AOne powerful strategy for dealing with noise and variability is to use ratiometric analysis techniques. In this strategy you put a normal sample and one exposed to the environmental stimulus on the same array. In doing that you can eliminate a lot of the variability between samples and arrays, since experimental and control samples are exposed to the exact same conditions on the same array. We also try to do everything we can to reduce the background and to be very consistent in the way that we do hybridizations. But it seems variability is inherent to microarray research.
QGiven this inherent variability, how do you analyze microarray data?
AWe have sophisticated error models that can normalize data for comparison and evaluate what constitutes a significant change in expression levels. But one of the most difficult issues facing microarray researchers now is that there’s not a standardized way for one lab to take their data and compare it to data generated by another lab under different conditions with a different set of reagents using different equipment and potentially different analysis techniques. The Microarray Gene Expression Database consortium, an international group of scientists from public and private sectors, is working on creating standards by which everyone will analyze their data. This will make the data far more valuable because it will vastly expand the amount of data available to individual researchers to compare to their own data in order to make observations of biological interest.
QIf you could make out a wish list for microarray technology advances or improvements over the next couple of years, what do you most want or need?
AThe first thing I would ask for is a consensus on how to analyze and compare the data. From a technology standpoint there are still a lot of areas with room for improvement, from labeling strategies to automation of hybridization to data capture. But these challenges have brought out some exceptionally creative thinking in technology development and I have every confidence that these technical issues will be resolved.