The complexity of biological systems is at times made worse by the diversity of ways in which they are described: the organic evolution of the science over many years has led to a myriad of conventions. This confusion is reflected by the in-silico representation of biological models, where many different computational paradigms and formalisms are used in a variety of software tools. The Systems Biology Markup Language (SBML) is an attempt to overcome this issue and aims to simplify the exchange of information by imposing a standardized way of representing models. The success of the idea is attested to by the fact that more than 110 software tools currently support SBML in one form or another. This work focuses on the translation of the Cyto-Sim simulation language (based on a discrete stochastic implementation of P systems) to SBML. We consider the issues both from the point of view of the employed software architecture and from that of the mapping between the features of the Cyto-Sim language and those of SBML.