There is increasing need for models and methods for analyzing multiple time course data with complex dependence patterns. Such data routinely arise in business and industry, physical and social science domains, and sampling scenarios conducted by national statistical agencies and private institutions. This invited session will present four talks discussing analysis for time course data with illustrations from diverse data domains, although the methods are  relevant and useful for statisticians and data scientists from many other domains as well. Since the data are large and complex, the talks will focus on computational feasibility as well. The topics will include 
(i) estimating large dense covariance matrices under changing financial market conditions from over 10,000 financial assets,
(ii) hierarchical Bayesian unit-level modeling of time courses from public-use micro survey data via efficient Gibbs samplers that can handle Gaussian, binary, or categorical responses,
(iii) Bayesian modeling and forecasting using R-Stan of volatility in irregularly spaced time series of returns from multiple financial assets on intra-day, transaction level data, and 
(iv) using efficient Markov chain Monte Carlo methods for dynamic modeling of multivariate integer-valued (count) models where dependence among the components is induced via the common environment which follows a Markovian evolution.