Since the discovery, almost 30 years ago, that cold atmospheric plasma (CAP) can inactivate bacteria, research in the field of plasma medicine has expanded into a wide range of possible biomedical applications. This includes research into various strategies to treat cancer, as CAP was shown later to both have the ability to kill cancer cells as well as to induce an immune response against them. While in vivo research is crucial to progress the field of plasma medicine in a clinical setting, in vitro research still lies at the basis of understanding the possible effects of plasma treatment on cells. In turn, modelling can provide fundamental insight into underlying mechanisms that are out of reach with experimental methods. In this thesis, we develop a 2D-axisymmetric fluid model, to investigate the transport phenomena that occur during the treatment of a well plate with a plasma jet, as commonly done for experimental in vitro research. With this model, we investigate the physical phenomena that occur in the studied system, and study examples where a lack of uniformity in plasma medicine research, both experimental and computational, hinders the direct comparison of results. In addition, we perform molecular dynamics simulations to gain insight into the actual, biological effects that plasma treatment may induce in treated (cancer) cells. Specifically, we focus on the immunotherapeutic potential of CAP treatment by modelling the effect of CAP-induced oxidation on HLA-E and HLA-Cw4, two protein ligands that are important for the crosstalk between cancer cells and natural killer (NK) cells. By comparing our simulations to experimental results, this data demonstrates the complex chemical and biological interactions between CAP and cancer cells, with regard to NK cell recognition, and provides an interesting starting point for further research. Taken together, the findings presented in this thesis provide a deeper insight into the in vitro treatment of cancer cells with a plasma jet on both a macroscopic and microscopic scale. We show various ways in which the chosen setup geometry can influence the treatment itself. It is important to keep these effects in mind as a source of variation in experiments, both when conducting experiments and when comparing them among each other. Moreover, the importance of benchmarking and comparison between computational results is underlined. Finally, we shed some light onto the immunotherapeutic potential of CAP treatment, and point toward possible paths for future research.