Codon Bias

The goal of the codon bias project is to explain the codon usage bias in DNA sequences.

There are biases in DNA that are translation independent and that are mainly observed at the di-nucleotide level. Detailed analyses of these biases indicate that they are most likely dependent on DNA-repair processes. This study led us to the investigate at the theoretical level whether biased DNA repair (which we call context-dependent DNA-repair) helps ensure that most mismatches are repaired such as to introduce silent mutations. Our data indicate that codon bias does in fact play an important role in the optimization of DNA repair. These studies also allow us to develop new tools to date mutations in DNA and derive new molecular clocks.

The distribution of codons is highly correlated in coding sequences. Our statistical data on codon distribution indicate that a selection pressure pushes towards reusing the same tRNA to translate successive occurrence of the same amino acid. A tRNA pairing index has been developed to measure the degree of reuse. Our data indicate that the observed effects are imposed by translation constraints. Experiments were designed to determine whether tRNA reusage does accelerate translation. Biochemical data indicate that it is the case, suggesting that tRNAs stick around the ribosome. Thus our statistical work enabled us to make predictions that we were able to test and confirm. Altogether, we show that tRNA diffusion is slower than translation, and that these differences in diffusion properties leave a signature on DNA.