Methods and apparatus are disclosed herein for encoding human readable text conveying a non-genetic message into nucleic acid sequences with a substantially reduced probability of biological impact and decoding such text from nucleic acid sequences. In one embodiment, each symbol of a symbol set of human readable symbols uniquely maps to a respective codon identifier. Mapping may ensure that each symbol will not map to a codon identifier that generates an amino acid residue which has a single-letter abbreviation that is the equivalent to the respective symbol. Synthetic nucleic acid sequences comprising such human readable text, and recombinant or synthetic cells comprising such sequences are provided, as well as methods of identifying cells, organisms, or samples containing such sequences.