#9 · Systems biology

Likelihood and first-passage frameworks extract kinetic parameters from stochastic data

The common thread is inference of kinetic parameters from explicit stochastic models, with estimated parameters mapped onto predicted outcomes. A preprint tutorial treats high-throughput single-molecule kinetic data through a first-passage time framework and maximum likelihood estimation, supplied with a GUI. The published agentBayes R package provides an asymptotically exact likelihood and Bayesian estimation for continuous-space agent-based reactant-catalyst-product models. A third preprint is less formal: it applies a simple statistical model to NarG sequence variants in soil nitrate reducers, relating variant reduction rates to environmental responses. This item shares the parameter-to-prediction logic rather than the stochastic formalism. What is new is asymptotically exact likelihood treatment for agent-based kinetic models, alongside an attempt to connect enzyme kinetic variation to environmental context.