Target frequencies and information content October 23

Given a query sequence, Q, of length m and a database, D, of length n, BLAST attempts to find all maximal segment pairs with statistically significant scores. Recall that a Maximal Segment Pair (MSP) is an ungapped local alignment whose score cannot be improved by extending or shortening the alignment. The data base is a series of concatenated amino acid sequences M1,M2, . . .Mj . . .. Figure 1 shows an MSP of length l between query sequence, Q and matching sequence, Mj . We use σ and τ , respectively, to designate the subsequences of Q and Mj that participate in the alignment.