Recent advances in single-cell RNA sequencing technologies enable deep insights into cellular development, gene regulation, and phenotypic diversity by measuring gene expression for thousands of cells in a single experiment. This results in high-throughput datasets and requires the development of new types of computational approaches to extract the useful and valuable underlying biological information of individual cells in heterogeneous biological populations. To addresses these approaches, in this paper, we introduce a deep learning technique to classify single cell types data from five primary Glioblastomas. We show that the deep learning method has the ability to correctly infer and classify cell type not used during the training process of the algorithm. Further, the deep learning method has the ability to identify the predictor variable Aquaporin 4 (AQP4), as the most important to make these predictions. Such computational approaches, as those presented in this study will enable researchers to better characterize the intratumoral heterogeneity in primary Glioblastoma.
