A community assessment of privacy preserving techniques for human genomes

Table 2 Results of Task 2.

	Teams	Top 1	Top 3	Top 5	Top 10	Top 15	Top 20	Top 30
Small (5000 SNVs)	UT Austin CMU	1 0.98	2.66 2.28	4.44 3.53	8.48 7.89	7.07 4.59	4.68 2.32	2.37 1.16
Large (100K SNVs)	UT Austin CMU	1 0.98	2.65 2.26	4.41 3.56	5.90 3.27	2.26 0.42	0.69 0.15	0.18 0.07

The table shows the average number of (1000 iterations) privacy-preserving SNV identification algorithms developed by the two participating teams. Both algorithms were trained using the small dataset consisting of 5000 SNVs, and then were tested on both small and large datasets, i.e., select top K (i.e., K = 1, 3, ..., 30) most significant SNVs.

ISSN: 1472-6947