In this paper we study generalizations of the
classical notion of solid codes and comma-free codes suggested by
DNA based computing. In particular we extend the study of coding
properties of languages that avoid certain undesirable Watson-Crick
bonds between the words. For an involution function we consider
involution codes: $\theta$-solid codes, $\theta$-overlap free and
$\theta$-join codes. $($ An involution function $\theta$ is such
that $\theta^2 = I$. An involution code refers to any of the
generalization of the classical notion of codes in which the
identity function is replaced by an involution function.$)$ We
formalize and investigate the properties of these codes and study
the closure and non-closure properties. Necessary condition for
$\theta$-solid codes to be maximal is also given. We also discuss
the properties of codes that are not $\theta$-comma free but can be
split into subcodes that are $\theta$-comma free. Note that if
$\theta$ is the identity function, the $\theta$-solid codes and
$\theta$-join codes respectively become the well known solid and
join codes respectively.
