Mining Large Heterogeneous Cancer Data Sets Using Boolean Implications

Abstract

Boolean implications (if-then rules) provide a conceptually simple, uniform and highly scalable way to find associations between pairs of random variables. In this paper, we describe their usage in mining associations from large, heterogeneous cancer data sets. Next, we illustrate how Boolean implications were used to discover a new causal association between a mutation and aberrant DNA hypermethylation in acute myeloid leukemia as well as the therapeutic implications of this discovery. We conclude with a brief description of how Boolean implications can be extracted from a given data set.