Siraela - In-Silico Metabolic Simulator with Application for Diabetes

In-silico modelling is essential in reducing adverse effects of clinical studies. Also, it reduces monetary costs when developing new drugs and devices. Diabetes is one of the diseases, where the in-silico modelling has established. Such an in-silico simulator mimics real metabolism and provides simulated responses to external treatments such as drugs and pump-delivered insulin. In this paper, we propose a novel approach, which addresses limits of the current metabolic simulators. First, a simulator is usually based on a set of differential equations, which are de-facto black-boxes for physicians. The proposed approach, Siraela, provides a language, which describes chemical reactions and bodily fluid equilibration in a manner lectured at medical faculties. Second, the current metabolic simulators are authority-accepted with single day scenario and few meals only. The proposed framework targets multiple weeks. Third, Siraela can be extended to other diseases, drugs and treatments, because its language is not bound to diabetes specifically. Fourth, Siraela requires readily-available equipment only. To demonstrate Siraela, we identified 133 profiles of real patients with average relative error 13.4\% and standard deviation 12.3\%