Carolin Loos introduces two novel approaches for theanalysis of single-cell data. Both approaches can be used to study cellularheterogeneity and therefore advance a holistic understanding of biologicalprocesses. The first method, ODE constrained mixture modeling, enables theidentification of subpopulation structures and sources of variability in single-cellsnapshot data. The second method estimates parameters of single-cell time-lapsedata using approximate Bayesian computation and is able to exploit the temporalcross-correlation of the data as well as lineage information.