Figure 1.

Segments consensus workflow. For clarity only three predictors are drawn. The orange elements represent predicted TM segments. (A) Scanning the results for the first segment. (B) Detecting overlapping segments and voting whether the group of overlapping segments should be added to the consensus result. (C) If the voting was positive (i. e. the SVM model predicts a TM segment - we assume that this is true in this case), a segment with averaged start and end positions was added to the consensus result (blue segment). (D) Masking the used segments and scanning for the next one.