A recent trend in computational methods for annotation of protein function is that many prediction tools are combined in complex workflows and pipelines to facilitate the analysis of feature combinations, for example, the entire repertoire of kinase-binding motifs in the human proteome.