-
Notifications
You must be signed in to change notification settings - Fork 27
antiSMASH Record Types
BiG-SCAPE 2 introduces the possibility to use any of the record types defined by antismash: (region, candidate cluster, protocluster and protocore) as the ‘working’ BGC record. For a full description of the region concept see the antiSMASH Documentation.
When protocluster or protocore record types are used, records within a 'chemical hybrid' candidate cluster or 'interleaved' candidate cluster are merged, to avoid duplication of records. This can be adjusted by updating the config.yml file (MERGED_CAND_CLUSTER_TYPES). Furthermore, in the BGC record visualization sections of the UI, the entire .gbk region will be displayed, and the relevant record domains will be shaded while the remaining domains will be semi-transparent (Fig. 2).
For every .gbk, BiG-SCAPE will try to extract the requested record type, if this is not present, BiG-SCAPE will try to extract the next higher level record type, i.e. if a proto_core feature is not present, BiG-SCAPE will look for a protocluster feature, and so on and so forth. The record type hierarchy is: region > cand_cluster > protocluster > proto_core.
Note: We advise new users to stick with either region or protocluster, as cand_cluster and proto_core will require more careful consideration of input data features and interpretation of results. Please refer to the antiSMASH Documentation for a full, comprehensive, description of how these records are generated and what features they include.
When making use of the record types that yield more than one record per region, two types of links can be made between records:
-
Sim-links, or similarity links, are the traditional BiG-SCAPE distances making up (solid-line) edges between nodes (records) in the sequence similarity network. -
Topo-links, or topological links, are formed between two nodes/records when these originate from the sameantiSMASH region, and are thus co-located in the chromosome. These links are represented by dashed lines in the BiG-SCAPE 2 sequence similarity network visualization (Fig. 2).