Real-time immune recognition

B-cells recognise a new antigen when it binds to a cognate, high-affinity receptor. In an acute primary infection, following initial recognition by high-affinity cells, a rapidly increasing antigen population could spread recognition to many low-affinity lineages. How do immune systems achieve specificity of recognition in real time? Here we show that B-cells can efficiently recognise antigens by a kinetic proofreading mechanism that is tuned to the complexity of the immune repertoire. The resulting proliferation-recognition dynamics of a primary infection can me mapped onto a generalised Luria-Delbrück process, akin to the dynamics of the classic fluctuation experiment. Our model predicts specific power law statistics of activated B-cells, which are in accordance with empirical data and explain clinical characteristics of acute infections, including the emergence of elite neutralisers and the effects of immune ageing. Thus, data from acute infections emerge as a new probe into the global architecture and functional principles of immune repertoires.