Forget Hard Drives And SSDs, Researchers Figured Out How To Store Data In DNA

Every cell in your body contains a bundle of nucleic acid—deoxyribonucleic acid, to be specific. More commonly known as DNA, this molecule contains all the genetic information that makes you a functional biological organism. Scientists have long been fascinated by DNA's incredible data storage density, but turning DNA into a general storage medium has been a vexing problem. That might be changing as researchers from the Rochester Institute of Technology (RIT) and the University of Minnesota have developed a biological circuit that uses DNA for both storage and processing.

DNA actually does many of the things that a traditional computer does. DNA, along with associated cellular mechanisms, supports sequencing, reading data, and writing data (protein synthesis). In a biological system, DNA bases (A, T, G, C) are grouped into threes. Cellular organelles "read" each trio of bases to assemble amino acids into proteins. This system gives DNA about 3-6 orders of magnitude more data density than the most advanced silicon-based systems. However, cold, lifeless technology is much more durable.

The key to turning DNA into a functional computing platform is a microfluidic integrated circuit (above). The lab-on-a-chip consists of a series of small channels studded with nanoscale sensors to separate, detect, and attract specific molecules. The researchers found they could represent numbers with varying concentrations of specially crafted "nicked DNA" snippets. This system can store and manipulate data as an artificial neural network.


This ultra-compact DNA chip is an impressive scientific achievement, but it's not going to replace your computer any time soon. The researchers say this system was designed in a decentralized manner—there's no silicon CPU overseeing a molecular computer, and the computing occurs "in-memory" to eliminate I/O bottlenecks. That comes with some drawbacks. For example, low concentrations of nicked DNA to represent variables close to zero yield few DNA strands in downstream reactions, which can lead to high error rates. Similarly, too many variables of high value can result in unintended chemical reactions that affect the quality of data.

The next step in this work is to develop a larger version of the microfluidic integrated circuit. Scaling up will help scientists better understand the functions and potential interactions of molecules. The team is hopeful this work can lead to useful molecular computing technology, but that's still a long way off. If it works out, these biological computers could be more sustainable than traditional electronic computers, which require rare and expensive materials.