Cambridge, UK, March 2013. My first days as a postdoc. Faced for the first time with data from ChIP–seq experiments, I was baffled. ChIP–seq (chromatin immunoprecipitation followed by sequencing) measures protein–DNA interactions genome-wide, but what do ChIP–seq counts reflect? They are supposed to reflect a binary measure, that is, the presence or absence of chromatin binding. So, was it meaningful to consider these counts as a continuous variable, or was I looking at measurement noise? My questions were addressed when I stumbled upon a 2006 paper by Amos Tanay.