Figure 6.

Number of predicted functions for proteins without annotations within CCS from hsa-dme-sce. For each subontology and similarity threshold the number of proteins without any annotation (olive), the number of proteins that receive new annotations (orange) and the total number of novel annotations are shown (yellow). Recall that a higher coherence threshold for CCS leads to less proteins being included in function predictions; thus, numbers generally decrease with higher thresholds.