Fig. 1

Domain Architecture (DA) examples. Both DAs have protein domains HAMP and MCPsignal, whereas only da01025 has CHASE3. When a sequence from da00101 is used as the query and retrieves a sequence from da01025, we classify the match as relevant (a “true positive”). Conversely, if a sequence from da01025 is the query and retrieves a sequence from da00101, the match does not fully recover the domain structures and therefore we classify it as irrelevant (a “false positive”)