Fig. 1.

Categories of the network clustering algorithm used in the protein complex prediction with PPI networks. The network clustering algorithms require as input either only a PPI network (methods in black color) or both on PPI network and biological information (methods in red color). Regardless of the input, the existing network clustering algorithms with applications to complex prediction can be divided into three categories, namely: node affinity-based, cluster quality-based, and network embedding-based methods. For each category, several examples are given and explained in this review. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)