Skip to main content
. 2019 May 23;15(5):e1006802. doi: 10.1371/journal.pcbi.1006802

Fig 6. Quality of solutions of UNCOVER on simulated data.

Fig 6

The fraction of genes in the planted (i.e., correct) solution found by the greedy algorithm and by the ILP approach are shown for different number of samples and the difference pn between the fraction p of samples with positive target and the fraction n of samples with negative target covered by the the correct solution.