Fig. 5

Prediction of secondary metabolites and their biological activities. A Gene cluster family (GCF) network of the 2334 identified biosynthetic gene clusters (BGCs). Each node represents one BGC. Only GCFs that contained more than one BGC are shown. BGCs that are not included in the 8 dominant BGC types are combined and shown in Others. Each cluster represents one GCF. The IDs of 33 GCFs that contain known BGCs from the MIBiG database are shown in the GCF network. B Prediction of biological activities for corresponding metabolite families of 33 known GCFs. The stacked columns represent the amounts and origins of BGCs in 33 GCFs. The heat map indicates the probabilities of different biological activities of 33 metabolite families. The metabolite families with high probabilities of antibacterial activity (> 80%) are shown on the stacked columns. The corresponding GCFs are highlighted in red