The ability to accurately predict the abundance of proteins from the expression of the corresponding genes has enormous potential for the advancement of biotechnological applications using metabolic engineering and synthetic biology approaches. Addressing this problem has been challenging because of the lag in methodological advances in quantifying protein abundances. Here, we reviewed and classified studies that investigated the relationship between gene expression and protein abundance in different experimental settings and cellular contexts. We focused on comparing and contrasting the findings based on different correlation-based measures, widely used with nominal or transformed transcriptomics and proteomics data. We also included studies that investigated and attempted to explain the observed correlations between gene expression and protein abundance by incorporating data on additional factors, such as translation rate, protein degradation, and post-transcriptional modifications, using various statistical and mechanistic modelling frameworks. Finally, we provided an overview of how the latest advances using data from single-cell analyses have contributed to addressing this pressing question. Our review offers a perspective about how the findings about the relationship between gene expression and protein abundance can propel biotechnological advances, particularly focusing on opportunities resulting from the availability of protein-constrained metabolic models and the complementary machine and deep learning models.