Skip to main content
. 2010 Jan 18;11(Suppl 1):S18. doi: 10.1186/1471-2105-11-S1-S18

Figure 4.

Figure 4

The performance comparison between AGL and BEAM (B) based on pure epistatic models. We generate 100 datasets for each model. Each dataset contains 400 sample (Nu = 200, Nd = 200) and 1000 SNPs. BEAM runs 5 × 106 Markov Chain Monte Carlo iterations which are 10 times of pairwise exhaustive search. The comparison shows that our method outperforms BEAM for these pure epistatic models.