I am maintaining a database of sequence files describing a collection of bacterial strains and plasmids that we have in our lab.

I am confronted to two sorts of issues:

• some sequence files contain errors
• other sequence files describe molecules that themselves contain variations (mutations, insertion or deletions) that occurred during the molecular cloning process.

I want to annotate the sequences to mark those errors and variations using the GenBank format. However, from the feature keys of the GenBank format described on the INSDC website, it is not clear to me which feature key should be used in each case.

To better explain my case, here below are what seems to me the most relevant feature keys, and my comment of the problem I have with them:

Feature key           variation
Definition            a related strain contains stable mutations from the same
                      gene (e.g., RFLPs, polymorphisms, etc.) which differ
                      from the presented sequence at this location (and
                      possibly others);
...
Comment               used to describe alleles, RFLP's,and other naturally occurring 
                      mutations and  polymorphisms; >>> variability arising as a result
                      of genetic manipulation (e.g. site directed mutagenesis) should 
                      be described with the misc_difference feature;
                      use /replace="" to annotate deletion, e.g. 
                      variation   4..5
                                 /replace=""  

According to the line marked >>> above, the feature key misc_difference should be used...

Feature Key           misc_difference
Definition            feature sequence is different from that presented 
                      in the entry and cannot be described by any other 
                      difference key (old_sequence, variation, or modified_base);
...
Comment               the misc_difference feature key should be used to 
                      describe variability that arises as a result of 
                      genetic manipulation (e.g. site directed mutagenesis);
                      >>> use /replace="" to annotate deletion, e.g.
                      misc_difference 412..433                                                     
                                      /replace=""  

But now, according to the example marked >>> above, it means the variation (here a deletion) is NOT in the considered entry... So how to indicate variations (mutation, insertion, deletion, errors) that are actually present in an entry ?