Introduction: After more than 10 years of use in research, induced pluripotent stem cell(iPSC) technologies are still under active development for clinical and disease modeling purposes. Directed differentiation protocols of iPSC to midbrain dopaminergic neurons enables the study human midbrain Dopamine(DA) neurons, the primary neuronal group affected in Parkinson's Disease(PD), in-vitro. Despite their increasing adoption and disease modeling potential in neuroscience research, establishing iPSC-derived in vitro DA neuron models presents challenges due to their high level of variability in their differentiation efficiency. The lack of robustness and consistency observed in those models limits their applicability. Single-cell transcriptomics is a powerful technology that can be used to investigate iPSC-derived models by dissecting their variability and characterize the heterogeneity of the single-cell states within these cultures. Objective: We use single-cell gene expression data of iPSC-derived DA neuron models to assess their cellular heterogeneity, in vivo biological relevance, and investigate novel use cases for PD research. The aim of this thesis is to understand how these factors influence current use cases of iPSC-based models in neuroscience research and to improve the existing model. Methods: We generate single-cell gene expression data sets of monolayer and organoids of Day 47-49 iPSC-derived midbrain DA cultures. We employ several computational techniques that leverage toxic and αSynuclein genetic stressor conditions, and demonstrate how in-silico methods can be used to perform exploratory analysis and infer disease signatures. Using genome editing, we investigate the impact of the familial PD mutations SNCA-A35T and SNCA-3x in monolayer and organoid DA neurons respectively. Through the use existing in vivo and in vitro data sets we try to understand the biological relevance of such in vitro models. We also developed a flexible computational framework for in-silico gating of gene sets in single-cell atlases. Results: Our iPSC-derived midbrain DA cultures show significant degrees of DA neuron cellular heterogeneity. Our computational method, \emph{scfind}, can use enriched gene sets identified in \emph{in vitro} cell types to discover biologically meaningful cell types using in vivo cell atlases. We conducted analyses which highlight that different \emph{in vitro} DA neuron cell types respond differently to genetic and environmental stimuli. Only a modest proportion of these DA neurons present in the culture shows transcriptomic convergence with \emph{in vivo} DA regions affected by PD. A similar subset of PD relevant DA neurons can be detected in both monolayer and organoid cultures. These DA neurons specifically, also exhibit high sensitivity to rotenone treatments and using in-silico techniques, we extract enriched molecular signatures of these cells that associate cholesterol biosynthesis and synaptic signaling. Through the use of cellular stressors, oxidative phosphorylation emerges as a converging downregulated pathway at a transcriptional level posing as the main driver of PD-associated pathogenesis in these models. Finally, we assess a co-culture organoid model as a sporadic PD model by elucidating the effect of culturing healthy cells with familial-PD cells within the same organoid. Conclusions: Human iPSC-based midbrain DA neuron in vitro models show great promise for PD research. However, at their current stage, iPSC technology and its differentiation protocols show high degrees of cellular heterogeneity which is detectable by single-cell gene expression. The use of single-cell gene expression assays provides insights into the culture composition, offers a robust way to investigate the molecular dysfunction phenotypes of PD, and should always complement bulk assays that aim to study the biology of iPSC-derived DA neurons. Furthermore, improvements in iPSC models should always be assessed in light of cell type heterogeneity and important in vivo genetic determinants.