French biotech DNA Script has a sense of timing. Barely finished the “Choose France” summit, it announced on Tuesday January 21 that it had received funding from the United States government to develop a technology for storing data in DNA.
In partnership with researchers from the Massachusetts Institute of Technology (MIT), Harvard, and Illumina – the American giant of genetic sequencing – she has four years to develop a machine capable of encoding a terabyte of data in a molecule – the equivalent of 250 films – in twenty-four hours at a maximum cost of 1,000 dollars (902 euros).
On paper, the idea is simple: it involves converting the data stored as 0 and 1 in a computer into a succession of nucleotides. These molecules designated by the letters A, T, C, and G make up the alphabet of DNA: thanks to the spectacular advances in molecular biology, it is now possible to read them – it is sequencing – but also to write – this is the synthesis. In 2017, Yaniv Erlich, professor of computer science at Columbia University had fun coding in DNA strands and then decoding The Arrival of a Train at La Ciotat, an 1895 film by the Lumiere brothers.
In the era of Big Data, this track is the most promising to meet future storage needs. Intelligence Advanced Research Projects Activity (IARPA) – an American intelligence agency – has invested a total of $ 48 million (43 million euros) in this “Moonshot” as we call across the Atlantic these projects intended to solve major problems using disruptive technology. A first envelope of 23 million dollars funds the consortium of which DNA Script is a part, and the remaining 25 million have been allocated to a second team which brings together Microsoft and the California biotech Twist Biosciences, champion of the DNA chip.
“Manage obsolescence”
According to IARPA, whose research is used by the National Security Agency (NSA), the CIA or the FBI, there is no time to lose. The data centers current ones occupy gigantic warehouses, consume megawatts of energy and cost billions of dollars throughout their life cycle: beyond a certain volume of information, that will become unacceptable, she notes.
“Building and operating an exabyte data warehouse costs more than $ 100 million. If you want a backup, it’s an extra $ 100 million with the challenge of syncing exabytes of data across two sites. “said David Markowitz, IARPA’s Molecular Data Storage (MIST) program manager.
–
