A New Informatics Method for Measuring Synonymous Codon Us Age Bias

Most of the current codon usage bias computational approaches are only suitable for the comparison of codon usage bias within a single genome. Here we introduce a new informatics method, referred to as synonymous codon usage order (SCUO), to measure synonymous codon usage bias. In this method, we used Shannon informational theory to describe the SCUO of each gene using a value ranging from 0 to 1, with larger values associated with greater codon bias. We compared our method with the codon adaptation index (CAI) method for measuring codon bias in Escherichia coli and Sacharomyces cerevisiae. We also studied the correlation between SCUO and CAI, and the relation of SCUO with gene length and gene function. Finally, we explored the correlation between SCUO and mRNA abundance in Sacharomyces cerevisiae using SAGE expression data. INTRODUCTION All amino acids except Met and Trp are coded by more than one codon. DNA sequence data from diverse organisms clearly show that synonymous codons for any amino acid are not used with equal frequency, even though choices among codons should be equivalent in terms of protein sequences (Grantham et al., 1980; Aota and Ikemura, 1986; Murray et al., 1989; Sharp et al., 1988; Shields et al., 1988; D’Onofrio et al., 1991). The relative frequency of synonymous codons varies with both the genes and the organisms. In Escherichia coli , genes of highly expressed proteins use codons corresponding to the most abundant tRNAs (Ikemura, 1985). Proteins expressed at low level use synonymo us codons in rough proportion to the abundance of the corresponding tRNA, and have weaker codon preference. In contrast, noncoding regions of E. coli DNA showed no pronounced preference for any codon. Recently, the constraint of tRNA contents on synonymous codon choice were confirmed in 18 different unicellular organisms (Kanaya et al., 1999).