Fig. 1
From: Assignment of structural domains in proteins using diffusion kernels on graphs
The overall workflow of KluDo. Each box represents a step in the KluDo pipeline. The arrows link algorithmically consecutive steps. S and M stand for single- and multi-domain, respectively. Also, M\(^{(={\textbf {1}})}\) and M\(^{(>{\textbf {1}})}\) show a single clustering and a set of candidate clusterings, respectively. The arrow from the step 5 (candidate clusterings) to single-domain output is drawn to show the cases for which no possible clustering exist and thus the protein structure is reconsidered as single-domain. The dashed arrow from the step 2 to the step 4 shows a special case where the number of domains is specified beforehand, so there is no need to use the single/multi-domain classifier