DNA STOR­AGE

SUC­CESS­FULLY PRO­GRAM­MING DNA COULD OPEN THE DOORS TO A BRAVE NEW ERA OF BI­O­LOG­I­CAL COM­PUT­ERS. ALEX REIS UN­RAV­ELS THE STRANDS

PC & Tech Authority - - CONTENTS -

For­get SSDs - DNA stor­age could be on its way

Our re­liance on gad­gets is cre­at­ing an un­prece­dented amount of data. Hu­mans pro­duce a staggering 16 zettabytes ev­ery year, which equates to 16 x 1021 bytes. And, last year, re­search group IDC pre­dicted that we’ll be cre­at­ing more than 160 zettabytes per year by 2025.

Hard drives and sil­i­con chips have served us well, but with all this data we’ll soon need more than they can of­fer. Faster. Big­ger. Bet­ter. Re­searchers are look­ing for the next big thing, and one of the po­ten­tial new ma­te­ri­als is rather sur­pris­ing: DNA.

BET­TER STOR­AGE

Re­searchers have known for a long time DNA can be used for data stor­age. In 1961, Richard Feyn­man was talk­ing about the po­ten­tial for sub-mi­cro­scopic com­put­ers, and the rst at­tempt at us­ing DNA for this pur­pose came in 1994 with Leonard Adle­man. Our ge­netic code is in many ways a per­fect match for a com­puter. After all, it stores the blue­print for mak­ing ev­ery liv­ing crea­ture on this planet.

As a bonus, it trans­mits from one gen­er­a­tion to the next with in­cred­i­ble re­li­a­bil­ity, with many of the genes to­day re­main­ing vir­tu­ally un­changed for count­less gen­er­a­tions. And, if you think DNA is frag­ile and del­i­cate, think again. DNA is in­cred­i­bly tough and long-last­ing; if kept un­der the right con­di­tions, it will stay in­tact for mil­lions of years.

It’s no big sur­prise, then, that re­searchers are try­ing to turn DNA into com­puter stor­age. DNA can be treated like a stan­dard stor­age de­vice: the bi­nary code comes from us­ing the bases thymine

(T), gua­nine (G), ade­nine (A) and cy­to­sine (C) to rep­re­sent 1s (T and G) and 0s (A and C). Re­searchers have al­ready squeezed an 1896 French movie, a com­puter virus, a $50 Ama­zon gift card, Shake­speare’s po­ems, a clip of Mar­tin Luther King’s “I have a dream” speech, and Wat­son and Crick’s work de­scrib­ing the struc­ture of DNA into DNA it­self. How­ever, “sav­ing” and “open­ing” les stored in DNA mem­ory doesn’t work ex­actly in a way we recog­nise. In fact, it’s a read­only process at the mo­ment, and the in­for­ma­tion has to be ac­cessed as a whole, not in sec­tions. If cur­rent com­put­ers were like that, you wouldn’t be able to save any new data and would have to open all the les in a folder at once. De­spite these dif cul­ties, in­ter­est in this eld has boomed over the past few years. In 2012, there were some of the rst at­tempts to go be­yond just cod­ing data into DNA. Re­searchers from Stan­ford Univer­sity suc­cess­fully wrote and rewrote one bit of data into bac­te­rial DNA. Their goal now is to in­crease from a sin­gle bit to eight bits – a byte – of pro­gram­mable ge­netic data stor­age.

In the in­ter­ven­ing years, re­searchers at the Univer­sity of Illi­nois took this a step fur­ther and en­coded the Wikipedia pages from six Amer­i­can uni­ver­si­ties, be­fore suc­cess­fully nd­ing and edit­ing parts of the text from three of those in­sti­tu­tions, within the DNA. In this case, the re­searchers “pass­word­pro­tected” each block of in­for­ma­tion with a speci c code, to make it eas­ier to ag up the sec­tions to nd.

The lat­est de­vel­op­ment comes from a col­lab­o­ra­tion be­tween uni­ver­si­ties in Italy, Swe­den and Ire­land, where re­searchers are tak­ing ad­van­tage of bac­te­ria and their small rings of DNA called plas­mids. Cru­cially, these micro­organ­isms “swap” plas­mids be­tween them­selves in a process known as con­ju­ga­tion.

The idea is to “save” data in plas­mids trapped in a speci c lo­ca­tion. To “open” these les, re­searchers send mo­bile bac­te­ria to visit their trapped coun­ter­parts. After con­ju­ga­tion, they re­turn car­ry­ing the de­sired chunks of data. “If bac­te­ria get within each other’s reach, in­for­ma­tion, in [the] form of DNA, can pass from a donor to a re­ceiver,” said Al­berto Giaretta, doc­toral stu­dent at Öre­bro Univer­sity in Swe­den and one of the au­thors of this study.

“Our idea is to build an ar­chive by en­cod­ing in­for­ma­tion in non-motile bac­te­ria [un­able to pro­pel them­selves, and there­fore im­mo­bile]. Later on, this

in­for­ma­tion can be read by motile bac­te­ria that, by us­ing a sort of GPS for bac­te­ria, [will be able to] move to­wards the ar­chive, read the in­for­ma­tion through con­ju­ga­tion and then de­liver such in­for­ma­tion to a third point.”

In keep­ing with tra­di­tion, the team gen­er­ated the DNA se­quence coded for the mes­sage “Hello World”, which was in­serted into a group of trapped bac­te­ria and suc­cess­fully re­trieved after con­ju­ga­tion from a group of motile bac­te­ria. “We used sev­eral known tech­niques, but in a dif­fer­ent way and for a dif­fer­ent pur­pose – a clever way of us­ing known molec­u­lar bi­ol­ogy tech­niques for a very dif­fer­ent ap­pli­ca­tion”, added Lee Cof­fey and Tri­ona Doo­ley-Cul­li­nane, re­searchers at the Water­ford In­sti­tute, Ire­land.

GROWABLE COM­PUT­ERS

While this is cer­tainly im­pres­sive, stor­age isn’t the only ap­pli­ca­tion for DNA in com­put­ers. In­cred­i­bly, re­searchers at Manch­ester Univer­sity have shown DNA can even be “taught” to per­form op­er­a­tions.

“Cur­rent com­put­ers work on the prin­ci­ple of read­ing a code (stored on the hard drive) and per­form­ing a com­mand (us­ing the mem­ory and pro­ces­sor),” ex­plained chemist An­drew Cur­rin, one of the au­thors in this study. “Our DNA com­puter has the same prin­ci­ple, ex­cept that our hard drive is the DNA se­quence and the pro­ces­sor is the en­zyme used to copy the DNA. You could eas­ily imag­ine DNA stor­age and DNA com­put­ers would work very well com­bined to­gether.”

What dis­tin­guishes DNA com­put­ers from our run-of-the-mill de­vices is that they can “grow”. Not fig­u­ra­tively, but lit­er­ally. As DNA per­forms a com­mand, it repli­cates it­self and dou­bles in ca­pac­ity. “Ev­ery­thing hap­pens in a tube. No liv­ing cells are used, and the DNA is en­tirely syn­thetic,” Cur­rin said. “The DNA code is recog­nised by a shorter piece of DNA, which then causes the rest of the DNA to be copied. Once the code is recog­nised, it can be specif­i­cally al­tered to make a new com­mand. This is done by a process called PCR [poly­merase chain re­ac­tion], a widely used tech­nique used to copy DNA.”

The con­se­quences of this ca­pac­ity in­crease are in­cred­i­ble. If you imag­ine a com­pu­ta­tional ques­tion as a maze, DNA com­put­ers take a com­pletely dif­fer­ent ap­proach to solve the prob­lem com­pared to stan­dard de­vices. “Stan­dard elec­tronic com­put­ers, when they come to a T-junc­tion, have to choose which path to take, whereas [our DNA com­puter] doesn’t need to choose, as it repli­cates it­self to fol­low both paths at the same time, thus find­ing the an­swer faster,” said Philip Day, reader in syn­thetic bi­ol­ogy and quan­ti­ta­tive ge­nomics at the Univer­sity of Manch­ester.

In other words, it’s like us­ing mil­lions of com­put­ers at the same time to solve the prob­lem. “In our DNA com­puter, each com­pu­ta­tion is rep­re­sented by a sin­gle DNA strand, which al­lows us to utilise many tril­lions of com­pu­ta­tions hap­pen­ing at the same time. This type of DNA-based com­puter can have huge ad­van­tages over con­ven­tional com­put­ers. We could have a com­puter more pow­er­ful than all com­put­ers in the world com­bined and fit it in a pocket,” ex­plains Kon­stantin Korovin, se­nior lec­turer at the Univer­sity of Manch­ester.

DNA CLOUDS

How­ever, there are still many chal­lenges to over­come. All this work re­lies on ef­fi­cient DNA se­quenc­ing which, al­though it has taken huge leaps since be­ing in­tro­duced in the late 1970s, is still bulky and ex­pen­sive. In ad­di­tion, the kind of prob­lems DNA com­put­ers can solve right now wouldn’t be of much use for post­ing a pic­ture on Face­book or writ­ing a Word doc.

“Cur­rently we have a proof-of­con­cept implementation, but we need to de­velop the tech­niques fur­ther to achieve the po­ten­tial. One of the tech­ni­cal chal­lenges is to make DNA com­pu­ta­tions re­li­able at a large scale and min­imise the num­ber of er­rors in com­pu­ta­tions,” says Korovin.

What makes this field ex­cit­ing as we ap­proach the next decade is that large com­pa­nies are start­ing to no­tice the po­ten­tial hid­den in­side DNA. Mi­crosoft has re­cently an­nounced its in­ter­est in adding DNA stor­age to its cloud sys­tem. “In­ter­est from tech­nol­ogy lead­ers such as Mi­crosoft, and re­search de­vel­op­ing at the pace which it cur­rently is, makes it likely that DNA stor­age will be a re­al­ity ver­sus a fan­tasy in the com­ing years,” said Cof­fey and Doo­ley-Cul­li­nane.

It will take some time for DNAbased com­put­ers to come into ex­is­tence – if they ever do – but there are many other po­ten­tial ap­pli­ca­tions. “We have some thoughts as to how DNA com­put­ers might be made avail­able, and one such idea is that these com­put­ers will be firstly ac­cessed through the cloud and used to work on larger com­pu­ta­tional prob­lems,” said the Univer­sity of Manch­ester’s Philip Day.

It’s not that dif­fi­cult to imag­ine DNA com­put­ers be­ing in­tro­duced into liv­ing cells to min­gle with ex­ist­ing bi­o­log­i­cal mech­a­nisms. Ex­tra­or­di­nary ex­am­ples on the way in­clude an in­tel­li­gent method to de­liver drugs only when needed or a more ac­cu­rate de­tec­tion of can­cer.

If sci­en­tists can crack DNA com­put­ers, it may not be long be­fore the lines be­tween nat­u­ral and ar­ti­fi­cial pro­gram­ming blur.

DNA isn’t just use­ful for data stor­age – in­cred­i­bly, it can also be taught to per­form tasks

Newspapers in English

Newspapers from Australia

© PressReader. All rights reserved.