Figure 2

Flowchart of the LocateP pipeline. Firstly, the possibility of being secreted by the Tat pathway was calculated by combining Tat-find v1.2 [91] and our Tat-specific HMMs (RR-HMM, CS-HMM). Bacteriocin-like proteins were identified using Bagel [149]. Secondly, Phobius [14], PrediSi [98], SignalP 3.0 [18] and TMHMM 2.0 [12] were combined to identify transmembrane regions. Those proteins without any predicted TM segments were considered intracellular, whereas those with TM segments were divided into multi-TM membrane proteins, N-anchored membrane proteins or secreted/released proteins (single N-terminal TM segment, possibly signal peptide), and C-anchored membrane proteins (signal peptide and single C-terminal TM segment). Thirdly, a sortase-substrate HMM [165] was used to distinguish LPxTG-type peptidoglycan-anchored proteins from C-anchored membrane proteins. Subsequently, signal peptidase type II (SPII) substrates were predicted by combining existing lipoprotein motif models [41, 157] and new lipoprotein HMMs. The remaining proteins were classified into the categories secreted/released or N-anchored membrane proteins. See Methods and additional file 1 for more details. Abbreviation: A-S = Anchored-Secreted; TMS = TransMembrane Segment; SP = Signal Peptide; C/N-TM = C/N-terminally transmembrane anchored; LPxTG = LPxTG cell-wall anchored.