Skip to main content
. 2013 Oct 15;14(Suppl 15):S18. doi: 10.1186/1471-2105-14-S15-S18

Figure 3.

Figure 3

The plots show the quality of the scoring function as measured by the average log likelihood of a unit of coevolution in our solutions versus the average log likelihood of a unit of coevolution in the reference matchings. Points are colored according to (a) recall and (b) relative gap size.