Precision in a rush: decision making in early fly development

Despite very limited time, organisms develop in reproducible ways. In the early stages of fly development positional information is read out in a few minutes to produce steep and precise gene expression patterns — the rough blueprint for future body parts. Motivated by recent live imaging experiments in fly embryos, I will discuss the speed-accuracy trade-offs. Traditional arguments based on fixed-time sampling of Bicoid concentration indicate that an accurate readout is not possible within the short times observed experimentally. I will compare fixed-time sampling strategies to decisions made on-the-fly, which are based on updating and comparing the likelihoods of being at an anterior or a posterior location. I will show that these more efficient schemes can complete reliable cell fate decisions even within the very short embryological timescales and discuss the the role of promoter architectures on the mean decision time.