E-Letter responses to:

p-forum: Zhen Lin, Art B. Owen, and Russ B. Altman GENETICS: Genomic Research and Human Subject Privacy Science 2004; 305: 183 [Summary] [Full text] [PDF]

E-Letters: Submit a response to this article

Published E-Letter responses:

Patient privacy and SNP detection through gene expression profiling

David J. States   (30 July 2004)

Patient privacy and SNP detection through gene expression profiling 30 July 2004
David J. States, Scientist University of Michigan Respond to this E-Letter: Re: Patient privacy and SNP detection through gene expression profiling	Modern gene expression profiling technology is potentially capable of revealing SNPs in expressed gene sequences. For example, the Affymetrix U133 plus 2.0 GeneChip analyzes the expression of 54,000 transcripts using an array composed 25 mers organized as 20 match-mismatch pairs per transcript. Effectively, the hybridization signal is sampling 27 megabases of genomic sequence. Given the average density of polymorphisms in the human genome, these GeneChip probes are likely to cover on the order of 20,000 SNPs. If anomalous patterns of hybridization permit the scoring of even 1% of these SNPs, then the gene expression data effectively contains a unique patient identifier. One approach to guarding patient confidentiality would be to report only processed gene expression profiles, rather than the raw data supporting those profiles. Such an approach would severely limit the ability of other scientists to assess the validity of conclusions or to reanalyze data with newer software tools. A more desirable approach would be to design probe sequences which avoid SNPs that are prevalent in the human population. This latter approach has the added advantage of improving the reliability of gene expression profiling by minimizing contamination of the signal by genetic variation in the population. David J. States, M.D., Ph.D., Professor of Human Genetics, Director of Bioinformatics, University of Michigan