AT A GLANCE
Assistant Professor of Environmental Medicine at University of Rochester’s Aab Institute of Biomedicine.
Received his PhD in microbiology and immunology, with special focus in molecular medicine and gene therapy, from the University of Rochester Medical School in 1999.
Avid Golfer. Competes in local amateur matches when he can get time away from his lab and the core facility.
QHow does the functional genomics center operate?
AThe functional genomics center, which we launched in January, offers genomics related services, including microarrays, sequencing, real-time quantitative RT-PCR, and proteomics, both to researchers within the university and on a fee-for-service basis to those at over thirty other institutions, as well as industry. With the microarray program, the core performs all wet lab work and queries our database to make sure that array data are within range for samples of that type. In addition to your raw data files we give you commercial and proprietary options for data analysis. We don’t want users to have to worry about the reproducibility associated with the technology.
QWhat types of microarrays do you use and in what combination?
AWe have one of the largest academic Affymetrix core facilities in the country. But now we are getting more heavily into printing our own arrays, using both oligonucleotide and cDNA approaches. People start with more broad experimental approaches using Affymetrix arrays, then refine the information they are looking for and we generate custom arrays to answer more specific questions. We also use a lot of Corning’s arrays. Our philosophy is that we’re going to find a technology that best suits our investigator’s needs. We’ve had requests for pig and dog arrays, and we’ve even done some work in chinchilla for a group in Buffalo.
QWhat kind of arrayer do you use?
AWe have a Virtek ChipWriter pro, a Virtek ChipReader, and an Affymetrix GeneChip system. We also have an ABI 7900 HT for real-time quantitative RT-PCR which is routinely used for array data validation studies or for investigators who need to query far fewer genes.
QWhat methods do you use to analyze microarray data?
AWe are not biostatisticians or bioinformaticists. We help people define the questions they want to ask with their microarray experiments and help them implement their ideas based on our experience. The problem with this scientific approach is it’s really new, and people get so enamored with the technology that they overlook the fundamentals of good experimental design. You can have great microarrays and reproducible sample prep, but unless you formulate your questions and the experimental design correctly, it is going to be very hard to make best use of your data.
QWhat is the biggest challenge you face in working with microarrays?
AThe biggest challenge is getting people at different institutions to standardize protocols so we can accurately compare array data from lab to lab. The other is in data analysis. If there were some sort of consensus about how to cleanse and normalize data, investigators would at least be able to compare data sets with colleagues at different institutions. Comparing across platforms is also a challenge, where you have cDNA vs. Affymetrix vs. spotted oligos.
QHow do you tackle these challenges?
AWe have developed various data analysis methodologies that allow investigators to contrast and compare their data sets. If you have two different cores it’s much harder. But as protocols become more uniform and the analysis applied to arrays becomes more standardized it will become less of a problem.
QWhat is your microarray wish list?
AIt would be great if we had an array where we could look at gene expression, protein expression, and metabolites at the same time. I would also like to see companies better define the “biological” sensitivity of their systems, so investigators will know if the gene expression changes are biologically relevant. Also, I would like to develop better arrays that can detect gene expression levels that fall below the sensitivity of the current array systems.