PrivAS is a tool to perform Genome-Wide Association studies (GWAS) using the Weighted-Sum Statistic (WSS) algorithm in a Privacy-Preserving environment. The underlying scenario takes into account three interacting parties:
- a Client, e.g. a genomic research unit, wanting to measure the association between an observed phenotype and regions of the genome
- a Reference Panel Provider (RPP) possessing genetic data for a Reference Panel, e.g. a priori healthy individuals of a carefully selected ancestry
- a Third-Party Server (TPS) with large computational capacities.
Our tool and its underlying implementation preserve both state-of-the-art performances and Privacy for all parties. Indeed, through a series of hashing and encryption mechanisms, we can assure that no genetic data from neither the Client nor the RPP are visible by the other parties involved. Furthermore, only the Client is able to view a decrypted version of the WSS results.