DNA Data Storage

Some ideas inspire awe, some ideas inspire the comment, "Damn, why didn't I think of that?"

A recent paper in Biotechnology Progress provoked the latter comment from me. Here Yachie et al. 2007 describe a living data storage system. The ability of DNA to encode information is well known; DNA sequences code for amino acid sequences of proteins. Combine that with the ability to construct custom DNA sequences and to clone specified DNA sequences into organisms, you've got a super data storage system.

Yachie et al. write, "Duplicated data encoded by different oligonucleotide sequences was inserted redundantly into multiple loci of the Bacillus subtilis genome. Multiple alignment of the bit data sequences decoded by B. subtilis genome sequences enabled the retrieval of stable and compact data without the need for template DNA, parity checks, or error-correcting algorithms."

Oh that's pretty cool. Say you want to save the information, "Mary had a little lamb." You could have nucleotide base triplets signify the letters of the alphabet, i.e. aaa = A, aac = B, aag = C and so on. Construct a DNA sequence reflecting your code for "Mary had a little lamb." Insert the oligionucleotide sequences redundantly into multiple locations in the non-coding regions of the B. subtilis genome. When you want to read your message at a later time, use primers to amplify the coded regions, then sequence and enjoy! The redundancy ensures that any errors (i.e. mutations) will be detected during multiple alignment software.

Oh and there's more. B. subtilis can be induced to form spores that can survive for millions of years. Compare that to DVDs and CDs. Seriously... cassette tapes were more reliable than CDs. Of course storing info this way is a bit expensive, but when the costs drop, perhaps we will use bacteria or viruses instead of silicon chips!

Photo by Yoshiaki Ohashi.