Fine-tuning of the genetic code itself

The genetic code — the mapping between codons and amino acids — appears optimized to minimize the effects of substitution mutations. Codons that differ by a single base tend to specify either the same amino acid or one from a related chemical group, softening the impact of transcription errors. This error-buffering property cannot have evolved by gradual modification because changing the codon-amino acid mapping would require simultaneously changing every protein produced by the cell — catastrophically disrupting cellular function. Therefore, this highly optimized code must have been set up from the beginning by an intelligence.