Stochastic models for single‐cell data: Current challenges and the way forward