Observed (column 4) and expected (column 5) numbers of detected non-canonical peptides compatible with translations according to each nuclear and mitochondrial vertebrate genetic codes. Predictions account for peptide length (mean length and standard deviation in columns 2 and 3), considering that translation of 60/64 (0.9375) codons is identical between these genetic codes. Results indicate strong biases against detection of peptides compatible with both genetic codes, showing that detected populations of peptides are specifically translated according to the mitochondrial vertebrate genetic code. This systematic bias excludes that detected non-canonical peptides have cytosolic origins.