This page displays the results of using SocialGene to search each MIBiG BGC against a database containing 343,381 RefSeq genomes (up to November 14, 2023), but only returning the best BGC (or partial BGC) hit within an organism available in a culture collection (up to 100 genomes). Protein searches were completed based on HMM-annotation similarity (HMM models from antiSMASH, AMRFinder, PFAM, Resfams, and TIGRFAM) and similarity links in the plot are the result of reciprocal best hits (RBH) between the proteins of the BGCs (using DIAMOND blastp --ultra-sensitive) Search parameters can be found here. It's difficult to create a search that works for every BGC. For this demo I tried settings that work well generally. However, multiple long genes like PKS/NRPS in BGC0000106 have a chance to break the search if consecutive PKS/NRPS are broken in the target sequence.

You have to refresh the page before selecting a new BGC. In some browsers you have to click the BGC selector a second time. If you want to contribute a fix, see: here.