Logic Regression is a new adaptive regression methodology that attempts to
construct predictors as Boolean combinations of (binary) covariates. In thi
s paper we use this algorithm to deal with single-nucleotide polymorphism (
SNP) sequence data. The predictors that are found are interpretable as risk
factors of the disease. Significance of these risk factors is assessed usi
ng techniques like cross-validation, permutation tests, and independent tes
t sets. These model selection techniques remain valid when data is dependen
t, as is the case for the family data used here. In our analysis of the Gen
etic Analysis Workshop 12 data we identify the exact locations of mutations
on gene I and gene 6 and a number of mutations on gene 2 that are associat
ed with the affected status, without selecting any false positives. (C) 200
1 Wiley-Liss, Inc.