Here is the summary of responses I received and those that were posted onto
the board regarding the question of prevalence of multimers of inserts and
vectors that was originally posted by Johan de Boer(jdboer at sol.UVic.CA).
All the responses indicated that dimers and other concatomers of the Vector
and inserts are possible but at a very low percentage of the overall clones.
You may have an increased likelihood of seeing these forms if you use RecA+
hosts such as MC1022. On the other hand use RecA- hosts such as DH5alpha you
will be selecting against the multimer forms. Paul Hengen wrote that he
could eliminate the doublet types by patching onto media containing high
levels of carbenicillin. Although this is a surprise for newcomers like me,
this phenomenon seem to have been well documented.