Amino acids in the cell differ in their abundance. Some of them are energetically expensive to make. tRNA species also differ in their abundance. Highly expressed genes (HEGs) should maximize the usage of cheap and abundant amino acids and use codons decoded by the most abundant tRNAs. Strong selection for optimizing translation elongation in HEGs, especially HEGs in rapidly replicating organisms such as Vibrio natriegens, results in codon adaptation and optimized amino acid usage. These HEGs are our guide in designing optimally translated mRNAs in mRNA medicine. Mutation tends to disrupt adaptation. How to quantify the effect of selection and mutation on codon usage bias? What codon usage indices have been used to measure the degree of codon adaptation? Why is the index of translation elongation (ITE), which takes background mutation bias into account, more preferable than the codon adaptation index, which ignores background mutation bias? Which codon usage index can predict protein abundance better than others? Does codon optimization of a gene really increase the protein production of the gene? Does this depend on translation initiation? Why is it mostly incorrect to use the dN/dS ratio as an indicator of positive selection? This chapter addresses these questions and many others.