A Fast Contiguous Sequential Pattern Mining Technique in DNA Data Sequences Using Position Information
SyedaFarzana Zerin, Byeong-Soo Jeong · IETE Technical Review · 2011
AbstractIn this paper, a position-based method is proposed for frequent contiguous DNA pattern mining by fast joining the same length patterns to generate next length ones as well as scanning the database only once. At first, we combine the position information of fixed-length patterns to generate a fixed-length spanning tree to mine frequent fixed-length contiguous patterns. Later on, by joining them based on their position information, we come up with a fast approach to mine all contiguous frequent patterns. After generating the fixed- length contiguous frequent DNA patterns, joining of the patterns of same length is done to produce the next length frequent patterns by comparing their sequence IDs and starting positions in the database. Besides, the spanning tree is constructed in such a way so that the frequent patterns are generated in sorted order. The sorted frequent patterns are then utilized for indexing the starting and ending position of the patterns with the first element as A, C, G, and T, res...