Figure 1
From: SS-Wrapper: a package of wrapper applications for similarity searches on Linux clusters
Effective search space vs. database size and query sequence length. Relationship between effective search space and database size with different query sequence lengths. The GenBank non-redundant protein database (nr) was split evenly according to the modified bucket algorithm in order to construct databases of a size 1/64, 1/32, 1/16, 1/8, 1/4, 1/2, or 1 of the entire nr database. Query sequences of varying lengths were randomly assembled using a Perl script. A BLAST search was then carried out for each query sequence against each database. The effective search space and database size was extracted from the BLAST results and plotted for each query sequence. The length of each query sequence is indicated next to the line which plots the relationship between effective search space and database size for that query.