Characterizing and Evaluating Cell Specialization Through the Gini Index of Gene Expression: A TCGA Normal Vs. Tumor Case Study

Background The Gini index, introduced by the Italian statistician and demographer Corrado Gini in the first decades of the 1900s, is commonly used as a measure of statistical dispersion to evaluate income inequality within a nation. However, it is a powerful and effective measure to characterize any sample distribution and evaluate how far it is from a uniform one. Methods In this work we used the Gini Index as an effective and reliable measurement of the specialization of cells, using it to evaluate and compare the specialization level of normal and tumor cells according to their gene expressions. Results It turned out that, on average, tumor cells tend to lose their specialization or, in other words, their capacity to be the cells they were intended to be due to cancer effects. This loss of specialization in tumor cells corresponds, in our analysis, to a lower Gini Index with respect to normal cells. This behavior was observed both at a single patient level comparing Gini Indexes of coupled samples (from the same patient) and at a global level comparing distributions of Gini Indexes in normal and tumor datasets. Discussion This work demonstrates that the Gini Index (GI) effectively captures the loss of transcriptional specialization in tumor cells compared to normal tissues, with statistically significant differences observed both within patients and across cancer types, despite some exceptions, such as KICH and THCA. Conclusion In conclusion we are confident that GI could be a valuable and effective parameter to evaluate cell specialization and could provide significant insights in the context of cancer studies.