<p>Clusters of Differentiation (CD) molecules are cell surface molecules that exhibit dynamic expression on immune cells. This plasticity characteristic of CD molecules is the basis for the high abundance and diversity of immune cell phenotypes. However, phenotypic associations of immune cells due to co-expression and mutual exclusion expression lead to some extent to a narrowing of the size of the immunophenotype repertoire. Therefore, a systematic exploration of the expression correlation including co-expression and mutual exclusion expression between CD molecules can help to reveal the co-phenotype and lost phenotypes during immunophenotyping analysis. In this study, using bioinformatics methods, we first confirmed that the conventional widely used Pearson and Spearman correlations are only suitable for co-expression but not for mutual exclusion expression analysis. Using bulk and single-cell RNA sequencing data, we systematically investigated the expression correlation and anti-correlation or mutual exclusion of 386 protein-encoding CD molecules in human CD4⁺ T cells, CD8⁺ T cells, B cells, NK cells, and monocytes. Highly correlated co-expression networks were identified in these cell types. Using flow cytometry, three CD molecules including CD58, CD63, and CD147 were experimentally validated to be significantly co-expressed in T cells. In particular, CD58 and CD147 showed a high correlation in either CD4⁺ or CD8⁺ T cells. Co-expression was also observed in T cells from patients with primary Sjögren’s syndrome (pSS) and was found to be positively correlated with the increased cytotoxic potential of T cells. In addition, the combination of CD58 and CD147 provided a more precise definition of the T-cell phenotype, which would be beneficial for improving the diagnostic efficiency of pSS.</p>