Intracellular calcium dynamics in spontaneously active cells such as neurons or astrocytes is an information-rich readout of the physiological state of the cell. Methods for deriving mechanistic information from biological time courses, as well as for extracting cellular activity time courses algorithmically from imaging data, have significantly advanced in recent years but been mostly applied to neuronal data. At the same time, the role of astrocytes, a type of glial brain cells, in enabling cognition and in psychiatric diseases is beginning to come into focus. In the present work, we analyze calcium dynamics in astrocytes from a transgenic mouse model of 22q11.2 deletion syndrome (DiGeorge syndrome), an inborn condition associated with psychiatric disorders and other abnormalities of development. Methods of calcium imaging, computer vision, and Bayesian inference are applied to compare normal and deletion-bearing cells. Inference of highest-likelihood molecular kinetic characteristics from the intracellular calcium time courses pinpoints a significant change in the activity of the sarcoendoplasmic reticulum calcium ATPase (SERCA). Applying a SERCA inhibitor to the normal cells reproduces the differences detected in the deletion-bearing cells. We conclude that Bayesian kinetic inference is a useful tool for mechanistic dissection of complex cellular phenotypes in neuropsychiatric glia research. Its application can allow rapid, rigorous formulation of specific hypotheses concerning the underlying molecular mechanisms, prioritization of experiments testing such hypotheses, and, in the future, individualized functional molecular diagnostics.