Skip to main content

Table 1 Profile discriminatory quality (PDQ) for profiles generated from searches of six databases Results are compared for 6 databases, the combined database with all the Sargasso Sea sequences (Combined-nr), the public databases at the time of deposition of the sequences (Curr-nr (April 2004)), the same databases a year later (Curr-nr (April 2005)) and three databases of similar size (i) the current databases and sections eai, eaj and eak of the Sargasso Sea resource (Combined_itok), (ii) the current databases and sequences from the Sargasso Sea of length 250 or greater (Combined_GT250) and (iii) a fragmentised version of the April 2005 databases (Curr-nr (April 2005) Frags). Profiles were generated for all 237 target sequences and profile discriminatory quality is described in the methods section. The size of each database in thousands of sequences is also shown. The table shows that profiles generated from databases with a high proportion of fragments have less discriminatory power.

From: An analysis of the Sargasso Sea resource and the consequences for database composition

DBs with Many Fragments DB Size PDQ DBs with Few Fragments DB Size PDQ
Combined-nr 1414 K 85.22 Curr-nr (April 2004) 783 K 93.69
Combined_itok 1025 K 86.17 Combined_GT250 1054 K 91.74
Curr-nr (April 2005) Frags 1006 K 89.1 Curr-nr (April 2005) 1006 K 93.36