Messages hidden in plain sight, in words, invisible characters, pixels, sound, printed dots, stacked transparencies, DNA and the bytes of files, and the analyses that find them.
DNA Stego
Hides a message in DNA as Clelland, Risca and Bancroft did in 1999: each symbol three bases by their key, the message between two PCR primers, the strand among fragments of other DNA (random here), found again by its primers. Their sequenced strand reads “JUNE6 INVASION:NORMANDY” by the key their patent prints whole; the key Nature printed lacks a row and misprints two codes.
Terms
About this tool
History
In June 1999 Catherine Taylor Clelland, Viviana Risca and Carter Bancroft described in Nature a DNA microdot, after the photographic microdots of German spies in the Second World War: a message in DNA, hidden among fragments of human DNA and dried onto a full stop; one survived the post on a letter. Their test message named the invasion of Normandy. Bancroft and Clelland’s patent of 2001 prints their key and primers whole.
How it works
The key gives each of 40 symbols, letters, digits, space, comma, full stop and colon, a codon of three bases, drawn with a random number generator until each had its own. The codons sit between two primers of 20 bases: 5′ forward primer, message, the reverse primer’s complement 3′. Among other DNA the strand cannot be seen; a recipient who knows the primers amplifies it by PCR. Here the other DNA is random fragments of 50 to 150 bases, the strand among them as itself or its other strand, and finding it means finding on one fragment, read either way, the forward primer and after it the reverse primer’s complement, up to three wrong bases allowed.
How it was checked
Fig. 1d’s sequence was read from the colours of its trace and agrees with the patent’s FIG. 1D. By the patent’s key its 69 bases read JUNE6 INVASION:NORMANDY, and the message written by the key gives them back base for base. The key Nature printed lacks the row F, P, Z, 6, gives E and O other codes and prints 0 as a second U; by it the strand reads JUNO? INVASI?N:N?RMANDY. The figure’s forward primer lacks its first T, which the patent explains: cloning polished off the 5′-terminal T. In the tests a message hidden twenty times among 300 fragments is found every time, either way round, and a primer site with one wrong base is found when one is allowed.
Limits
Nothing here is chemistry: the other DNA is random letters, not a genome, and finding is matching text, not PCR, which tolerates mismatches by its own rules and can amplify stray sequences. New primers follow simple rules of the tool’s own, not the paper’s choice by melting temperature and folding. The key is a simple substitution: the paper puts the security in the hiding and the primers, not the cipher. The patent’s other uses, dot arrays to mark objects and messages in living cells, are left out.
- Catherine Taylor Clelland, Viviana Risca and Carter Bancroft, “Hiding messages in DNA microdots”, Nature 399 (10 June 1999), pp. 533–534
- F. Carter Bancroft and Catherine Clelland, US Patent 6,312,911 B1, “DNA-based steganography” (2001): FIG. 1B, the key; FIG. 1D, the sequence; cols. 11–12, the primers