Fig 4. Gene tree and representative protein domain architecture of 73 B. glabrata FReDs.

A. Maximum likelihood tree with 1000 bootstrap tests was constructed using full length representative protein sequences of 73 FReD genes in B. glabrata. Nodes with bootstrap support of 75 or higher are marked with black squares. The gene IDs of the 73 FReDs were followed with number of IgSF domains and the best hit with any published FREP genes (BLASTp hit with at least 150 aa aligned with ≥ 90% identity was considered as the same FREP; or considered a “FREP-like” gene if identity was 85% ~ 90%; identity <85% was not assigned to any published FREP). Any FReD genes with best hit to FREPs from the genome paper [32] are labeled with a red dot at the corresponding gene ID. Complete FREP or FReD genes fitting the criteria in this study were highlighted in blue text. B. Domain architectures were predicted using InterProScan and lineage specific HMM models. “Peptide sequences” represent multiple peptide sequences extracted from a proteomics study [49]. Colored small horizontal bars above or below some protein domain structures highlight the matched location of peptide sequences with binding affinity to S. mansoni sporocyst membrane-enriched and larval transformation proteins, from either schistosome-susceptible NMRI strain (SUS in green) or -resistant BS-90 strain (RES in red) of B. glabrata. The dashed gray line was manually added to distinguish most “sFReD clades” from “FREP clades”.