The VanBUG team occasionally forwards events of interest to the local
bioinformatics community. The upcoming event is posted on the
Bioinformatics Event Calendar at:http://www.vanbug.org/calendar/
Seminar Talk on Genomic Data Privacy
Ph.D. candidate, Berger Lab, MIT
Friday March 13 , 2015, 10:30am to 11:30am.
SFU Burnaby Campus, Applied Science Building, Room 9896
This talk is sponsored by the NSERC Frontiers in Discovery Program on the Cancer Genome Collaboratory
Title: One Size Doesn’t Fit All: Measuring Individual Privacy in Aggregate Genomic Data
The past decade has seen an explosive growth in the amount of genomic data available. Though this data is immensely valuable to biomedical researchers, its sensitive nature has led to privacy concerns. Recent work has shown that even aggregate genomic data (such as regression coefficients, p-values and count data) can leak private information about individuals involved in a genomic study.
After giving an overview of some of the major privacy issues in the genomics community, I will present a model-based measure, PrivMAF, which provides provable privacy guarantees for aggregate data (namely minor allele frequencies) obtained from genomic studies. Unlike many previous measures, PrivMAF gives a measure of privacy for each individual in a study and not just an average measure. These individual measures can then be combined to measure the worst case privacy loss in the study. PrivMAF also allows us to measure how perturbing the data, either by adding noise or binning, affects privacy. I also present results from applying PrivMAF to genotype data from the Wellcome Trust Case Control Consortium, providing a more nuanced understanding of the privacy risks involved in an actual genomic study.
Sean is a fourth year PhD student in the Berger lab at MIT. Before that he received his B.S. from the University of Texas at Austin. Currently he spends most of his time working on various computational problems related to biomedical privacy. He also occasionally plays around with using machine learning to solve interesting biological problems.