All tumors are characterized by intratumor heterogeneity at varying degrees. Cancer stem cells have been put forward to be an essential element that promotes heterogeneity. Myxoid liposarcoma, which is a lipogenic cancer that develops in deep soft connective tissues, is characterized by intermediate intratumor heterogeneity. Despite recent therapeutic advances, the post-treatment recurrence rate remains relatively high. Identifying sub-populations of myxoid liposarcoma tumors can help in characterizing their molecular signatures and tumorigenic capabilities leading to developing better therapeutics. Single-cell transcriptomic approaches can highlight deviations in gene expression patterns among different subpopulations within the tumor. In this study, a multi-algorithmic pipeline was developed to make a fast, simple and efficient process for characterizing cellular sub-populations of cancer cells and gain insight about the molecular signature of the cancer stem sub-population. This pipeline consists of four successive steps, read counts’ pre-processing, cellular clustering and pseudotemporal ordering, defining differential expressed genes and defining biomarker genes. The results showed a harmonic integration between the algorithms that constitute the backbone of the proposed pipeline leading to a reduction in the limitations of some of these algorithms. The outcome of this study is a panel of 33 genes nominated as possible biomarkers for stemness and aggressiveness. To optimize and validate these biomarker candidates, further investigations are required. Moreover, additional functional coupling analysis is necessary to nominate biomarkers for each of the sub-populations based on the defined differential expressed genes.