An Explanation of the SEED Evidence Codes

Within the SEED, we use evidence codes to reflect significant factors that go into making assignments of function. Some of these codes are computed and reflect information that we consider particularly useful. Others are used to reflect experimental evidence of function.

icw(n): in cluster with

This code indicates that the PEG occurs in a cluster with n other genes from the same subsystem (very strong evidence). There may be several of these for a PEG (up to one for each subsystem the PEG occurs in).

isu: in subsystem unique -- the only entry in a subsystem cell

This code indicates that the PEG occurs in a subsystem, and it is the only PEG for that genome that has been assigned the functional role (i.e., the cell in the spreadsheet contains a single entry). This means that, if you wish to change an annotation, you should discuss it with the owner of the subsystem.

idu(n): in subsystem duplicates

This code indicates that the PEG occurs in a subsystem, but it is in a cell of the spreadsheet containing duplicates (and it is not clustered with other genes connected to the same subsystem). In this case, you may make a change without notifying the owner of the subsystem, since you are probably disambiguating the situation to his benefit.

ff: in FIGfam

This code indicates that the protein-encoding gene is included in a FIGfam.

cwn: Clustered with Nonhypothetical

This code indicates that the protein-encoding gene is functionally coupled to at least one other protein that has been assigned a function that is considered "nonhypothetical". The functional coupling score must be 5 or more for this code to apply. This means that the gene co-occurs on the chromosome in at least 5 instances of genomes that are not close strains with another gene that is considered nonhypothetical.

cwh: Clustered with Hypothetical

This code indicates that the protein-encoding gene is functionally coupled to at least one other protein that has been assigned a function, but none that is considered "nonhypothetical". The functional coupling score must be 5 or more for this code to apply. This means that the gene co-occurs on the chromosome in at least 5 instances of genomes that are not close strains with at least one other gene, but none that is considered nonhypothetical.

dlit: Direct Literature References to the Gene Exist

This code is used to indicate that at least one paper (that is a "nongenome paper" in the sense that it does not reference hundreds of genes) is associated with this gene.

ilit: Indirect Literature References to the Gene Exist

This code is used to indicate that at least one paper (that is a "nongenome paper" in the sense that it does not reference hundreds of genes) is associated with a gene assigned the same functional role, but none to this gene itself (as far as we know).