I have investigated a few topics in stochastic simulations for chemical kinetics. I have been working on efficient algorithms for exact simulations. I have some adaptations of Gillespie's direct method and Gibson and Bruck's next reaction method that have better computation complexity and better performance than previous work.
As random number generation is important in stochastic simulations, I have implemented a package with uniform and non-uniform generators. It has quite a few of the recently developed algorithms. I've worked on improving the performance of Poisson generators (useful in tau-leaping) and have developed some new discrete, finite generators (useful in the direct method).
Previously I looked at concurrent algorithms for tau-leaping. Although concurrent algorithms may be formulated, they are not practical for generating a suite of trajectories on multi-core computers or networks of computers.