Figure 4

Hierarchical clustering of GMPA score criterion significant nickel associated proteins. Potential nickel proteins clustered based on co-occurrence in fractions using Ward's method. The grid at top contains excerpts from the data set selected to give a rough sense of protein co-occurrence within fractions and its effect on the resulting clustering. The numbered boxes at the bottom indicate the partitioning of the overall clustering into self-contained clusters as determined by cutreeHybrid (indicated by colors in supplementary material clusterings). Clusters containing known Ni-proteins are highlighted at the bottom illustrating the clustering together of subunits of known nickel-proteins and clustering apart of distinct Ni-proteins (SHI: soluble hydrogenase I PF0891-PF0894, SHII: soluble hydrogenase II PF1329-PF1332).