A new development in genomic research, the Genomic Tokenizer (GT), has been introduced to enhance DNA sequence analysis through a biology-driven tokenization method. This approach aligns with the central dogma of molecular biology by utilizing codons—three-letter sequences that represent amino acids. The Genomic Tokenizer incorporates start codons, synonymous codons, and stop codons into the tokenizer interface of the HuggingFace transformer package. This enables it to manage shifts in reading frames caused by nucleotide additions or deletions. Additionally, the tokenizer preserves biological nuances inherent in genetic variations during the tokenization process. The vocabulary of the Genomic Tokenizer includes all possible codons, with synonymous codons assigned the same IDs to improve efficiency.