Case-control data.

For case- control data, Treepeeling calculates the likelihood of the phenotypes conditional on two penetrance parameters, the probability of a chromosome with a mutant allele to be carried by a case individual in the sample and the same probability for a wildtype allele. The algorithm can either be run for a known set of penetrance parameters, or alternatively it can average over a grid of penetrance values. As the penetrance of a single locus in a pre-ascertained sample of cases and controls can differ from penetrance parameters ascertained through other means, it is usually advisable to integrate over the space of penetrances. The algorithm performs this analysis on a grid of penetrances for the wildtype and the mutant allele, the density of that grid must be decided on by the user. The number of gridpoints are then evenly distributed by the program. If the penetrance parameters are unknown, we suggest using the maximal gridsize of 20x20 points.