Two groups can share the same average and still look very different at the tails. Quantile regression asks how a covariate shifts the 10th, 50th, or 90th percentile of an outcome, with no need to assume a shape for the error distribution. This matters most in health data, where the people at the extremes are often the ones we care about.
This idea sits at the center of the lab. Each of the other areas takes quantile regression into a new setting: survival times, very many covariates, data spread over space, and large health studies.