Figure 3 Data forwarding with direct cluster head to cluster head communication as proposed in the algorithm.
Our clustering algorithm avoids this weakness by defining the direct cluster head to cluster head connectivity as shown in Figure 3.
We used three algorithms for clustering states in neural network connectivity dynamics: direct clustering of the functional network with the k-means algorithm, modularity-based and topological based clustering.
An important shared feature is the variation in brightness, which is both present in dimension 1 of the direct cluster analysis, and in the perceptual space depicting the spliced stimuli (from the same 19 clusters).
Knowledge is discovered from data, obtained using the SOM generalization aptitude and taking advantage of the well-known SOM abilities to discover natural data grouping when compared with direct clustering.
Whereas the 2-D clusters were constructed through a "direct" clustering of the compounds being considered, the 3-D cluster construction involved an "indirect" clustering of the compounds, meaning that their multiple conformers were clustered first, then the conformer identifiers were converted to their corresponding compound identifiers (i.e., CIDs).
Another reason not applying a direct clustering method for the automatic selection of three representative clusters - two on the edges, one in the centre - is that the trained distributions for each cluster would intersect and share data points, leading to false-positive detections.
In a direct clustering, the desired k-way clustering solution is computed by simultaneously finding all k clusters.
Hierarchical clustering was performed with the vcluster program from the Cluto clustering toolkit v.2.1.1 (http://glaros.dtc.umn.edu/gkhome/views/cluto/). The selected algorithm is based on the application of a direct clustering technique.
The chart in Figure S6 shows that our results described the data better than the direct clustering method (COGO: CV = 0.27 versus direct clustering: CV = 1.92).
A comparative study was provided to compare our method to the direct clustering method.
