Heterogeneity among patients commonly exists in clinical studies and leads to challenges in precision medicine research. It is widely accepted that there exist various subtypes in the population, and they are distinct from each other. Therefore, identifying the subtypes to apply the appropriate treatments is a vital step to precision medicine. The mixture model is a classical statistical model to cluster the heterogeneous population into homogeneous subpopulations. However, for highly heterogeneous population with multiple components, its parameter estimation and clustering results may be ambiguous due to bad local maxima. For subtyping purpose, we work on the finite mixture of regression models with concomitant variable to quantify the mixing probabilities and propose a novel statistical method to identify the components in the mixture sequentially.