Watermarking DNA

It's not cryptography -- despite the name -- but it's interesting:

DNA-based watermarks using the DNA-Crypt algorithm

Background

The aim of this paper is to demonstrate the application of watermarks based on DNA sequences to identify the unauthorized use of genetically modified organisms (GMOs) protected by patents. Predicted mutations in the genome can be corrected by the DNA-Crypt program leaving the encrypted information intact. Existing DNA cryptographic and steganographic algorithms use synthetic DNA sequences to store binary information however, although these sequences can be used for authentication, they may change the target DNA sequence when introduced into living organisms.

Results

The DNA-Crypt algorithm and image steganography are based on the same watermark-hiding principle, namely using the least significant base in case of DNA-Crypt and the least significant bit in case of the image steganography. It can be combined with binary encryption algorithms like AES, RSA or Blowfish. DNA-Crypt is able to correct mutations in the target DNA with several mutation correction codes such as the Hamming-code or the WDH-code. Mutations which can occur infrequently may destroy the encrypted information, however an integrated fuzzy controller decides on a set of heuristics based on three input dimensions, and recommends whether or not to use a correction code. These three input dimensions are the length of the sequence, the individual mutation rate and the stability over time, which is represented by the number of generations. In silico experiments using the Ypt7 in Saccharomyces cerevisiae shows that the DNA watermarks produced by DNA-Crypt do not alter the translation of mRNA into protein.

Conclusions

The program is able to store watermarks in living organisms and can maintain the original information by correcting mutations itself. Pairwise or multiple sequence alignments show that DNA-Crypt produces few mismatches between the sequences similar to all steganographic algorithms.

Paper here.

Posted on June 8, 2007 at 11:47 AM • 14 Comments

Comments

DanielJune 8, 2007 9:37 PM

Well, there's a "patent code" in an engineered prokaryote in _Memory_ by Lois McMaster Bujold...

gregJune 9, 2007 11:05 AM

I predict that many Gene patents will be ruled invalide. (Not all) soon. Also a lot of countries are writing in extra exceptions for use of unlicenced patents... So this my not be very economic.

Thomas DamgaardJune 10, 2007 1:08 PM

Isn't the point of patents that the inventor (or patent holder) *gives information about the invention to the public*; in exchange he/she gets a temporary monopoly?
Then why would they want to "encrypt" the genes?

StudentJune 11, 2007 3:28 AM

Is it just me that finds it rather disturbing that somebody is allowed to patent living organisms in the states?

I wonder how long it takes before the first patented human shows up. And if (s)he will have any rights to reproduce...

Mr. MikeJune 11, 2007 8:27 AM

What's next? DRM on Evolution (tm). I wonder if they are counting on Intelligent Design. :-P

(couldn't resist the puns)

UNTERJune 11, 2007 12:22 PM

@Damgaard:

The stated goal is not to encrypt the gene, but to encrypt within the gene a water mark, making it possible to differentiate a "naturally occurring" version of a gene from an equivalent patented version. It's about as sensible as the DRM version, blocking normal people from using an unauthorized version of the gene (since they will ostensibly just propagate the plant), yet do nothing against commercial piraters who will simply insert the gene sequence without the watermark in the lab.

Always going against the little guy.

UNTERJune 11, 2007 12:25 PM

On top of that, if there exists a "natural version", wouldn't that be prior art? And if nature evolves an equivalent version, is nature going to be sued for patent violation? How can you distinguish between "prior art" and "later art" in genes - is nature going to be required to publish in journals now to keep her rights?

Isn't patenting genes about as absurd as logically possible? How about patenting frequencies?

KevinJune 11, 2007 7:07 PM

With a genetically engineered human, if they're covered by patents, I guess they could be liable for patent infringment if they reproduced before the patent expired. Is it just coincidence that patents expire very close to the amount of time it takes an embryo to become an adult? Or did They(tm) plan it that way?

On a more serious note, what is covered by these patents is the mechanism of (say) growing corn with vitamin A in it by means of inserting such and such a gene. Patents expire, this sort of stuff does require heavy R&D, I don't have much problem with it. It's patents for little silly obvious stuff that bothers me.

guvn'rJune 12, 2007 6:36 AM

oh great, how when the GMO agricrop jumps the fence and runs wild the neighboring farmers aren't just victims of genetic pollution they get sued for IP theft!

@Kevin, problem with, say, Vit A enhanced corn is when it gets loose before the side effects (program flaws, analogous to software bugs) are discovered.

Base concept is interesting, but raises yet another concern about potential side effects, mutation paths, etc.

Tinkering with genetics is getting really close to trespassing on God's turf. Is that something we want fallible man to be doing, especially solely for economic gain?

X the UnknownJune 13, 2007 10:10 AM

@UNTER: "How can you distinguish between "prior art" and "later art" in genes"

Presumably with a Genetic WaterMark. I imagine that the argument would be very similar to the one for fingerprint-ID: "The chances of an otherwise-useless genetic sequence of this length being generated randomly is one in a gazillion..."

Leave a comment

Allowed HTML: <a href="URL"> • <em> <cite> <i> • <strong> <b> • <sub> <sup> • <ul> <ol> <li> • <blockquote> <pre>

Photo of Bruce Schneier by Per Ervland.

Schneier on Security is a personal website. Opinions expressed are not necessarily those of Co3 Systems, Inc..