Efficient and Accurate Discovery of Patterns in Sequence Datasets

Existing sequence mining algorithms mostly focus on mining for subsequences. However, a large class of applications, such as biological DNA and protein motif mining, require efficient mining of “approximate” patterns that are contiguous. The few existing algorithms that can be applied to find such contiguous approximate pattern mining have drawbacks like poor scalability, lack of guarantees in finding the pattern, and difficulty in adapting to other applications. In this paper, we present a new algorithm called FLAME (FLexible and Accurate Motif Detector). FLAME is a flexible suffix tree based algorithm that can be used to find frequent patterns with a variety of definitions of motif (pattern) models. It is also accurate, as it always find the pattern if it exists. Using both real and synthetic datasets, we demonstrate that FLAME is fast, scalable, and outperforms existing algorithms on a variety of performance metrics. Using FLAME, it is now possible to mine datasets that would have been prohibitively difficult with existing tools.