Department of Computer Engineering
CS 590 SEMINAR
Kinship Inference from Public Genomic Databases and its Countermeasures
Computer Engineering Department
With the advent of rapid and low cost sequencing technologies, the applications that utilize genetic data are growing steadily. Individuals share their genetic data in order to contribute to genetic research studies or to get personalized services. However, genetic data bear some of the most sensitive information about an individual. This kind of sensitive information, if passed onto wrong hands, may pose security threats. This work aims at identifying risks associated with public access databanks that store genetic data. In this project, using openSNP.org database, we show that it is possible to infer familial relationships up to third degree from this database in which the genotype information is anonymized. We achieve this through the use of clustering methods and appropriate kinship estimators. To alleviate these risks, we also aim at developing techniques that enable the controlled release of genetic data with minimal security concerns on familial relationship inference.
DATE: 21 March, 2016, Monday @ 16:50