GMOs and junk science

Financial Mirror (Cyprus) - - FRONT PAGE -

In to­day’s media land­scape, where un­founded opin­ions, hype, and ru­mours are rife, the sci­en­tific method – the means by which we de­ter­mine, based on em­pir­i­cal and mea­sur­able ev­i­dence, what is true – should serve as a touch­stone of re­al­ity. Science en­ables us to gauge what we think we know and to iden­tify what we do not. Most im­por­tant, it dis­cred­its false claims made for per­sonal or po­lit­i­cal rea­sons – at least it should.

But sci­en­tists oc­ca­sion­ally “go rogue,” for­sak­ing the sci­en­tific method – of­ten for no­to­ri­ety or eco­nomic gain – to pro­duce pro­pa­ganda and to sow fear in a public that lacks ex­per­tise but is hun­gry for in­for­ma­tion. This abuse of sci­en­tific au­thor­ity is es­pe­cially wide­spread in the “or­ganic” and “nat­u­ral” food in­dus­tries, which cap­i­talise on peo­ple’s fear of syn­thetic or “un­nat­u­ral” prod­ucts.

A re­cent ex­am­ple is the In­dian-Amer­i­can sci­en­tist V.A. Shiva Ayyadu­rai, who, with Prab­hakar Deonikar, pub­lished the muchridiculed pa­per “Do GMOs Ac­cu­mu­late Formalde­hyde and Dis­rupt Molec­u­lar Sys­tems Equi­lib­ria? Sys­tems Bi­ol­ogy May Pro­vide An­swers.” (“GMOs” are “ge­net­i­cally mod­i­fied or­gan­isms,” it­self a mis­lead­ing and of­ten un­fairly stig­ma­tised non-cat­e­gory, cir­cum­scrib­ing a uni­verse of or­gan­isms mod­i­fied with the most mod­ern and pre­cise tech­niques of ge­netic en­gi­neer­ing.)

Although the ar­ti­cle sup­pos­edly passed the peer-re­view process, a key com­po­nent of le­git­i­mate science, it ap­peared in a low­im­pact “pay-for-play” jour­nal, Agri­cul­tural Sciences, which is pro­duced by a “preda­tory” pub­lisher. Within days of pub­li­ca­tion, an­tibiotech­nol­ogy or­gan­i­sa­tions like the Or­ganic Con­sumers As­so­ci­a­tion and GMO In­side were re­port­ing on Ayyadu­rai’s “find­ings” with fright­en­ing head­lines ac­com­pa­nied by scary graph­ics.

But the prob­lems with Ayyadu­rai’s pa­per are le­gion. Its ti­tle alone is enough to show that some­thing is amiss. If you think that GMOs might “ac­cu­mu­late formalde­hyde” – a chem­i­cal that is prob­a­bly car­cino­genic at high lev­els but is present in most liv­ing cells and found widely in our en­vi­ron­ment – the ob­vi­ous re­sponse would be to mea­sure its lev­els in the or­gan­isms. Ayyadu­rai, how­ever, chose to make guesses based on mod­el­ing via “sys­tems bi­ol­ogy.”

“Sys­tems bi­ol­ogy” en­ables only a pre­dic­tion, not an ex­per­i­men­tal con­clu­sion. Rather than ac­tu­ally test­ing the lev­els of any chem­i­cals in plants, Ayyadu­rai plugged data into a com­puter al­go­rithm to pre­dict the lev­els of two chem­i­cals, formalde­hyde and glu­tathione. This is akin to a me­te­o­rol­o­gist pre­dict­ing from his mod­els that it will be sunny all day, in­stead of look­ing out win­dow to see whether rain is fall­ing.

To be sure, as Kevin Folta, the head of the hor­ti­cul­tural sciences depart­ment at the Univer­sity of Florida, ex­plained, sys­tems bi­ol­ogy can be a use­ful ap­proach if em­ployed prop­erly. As he put it, sys­tems bi­ol­ogy “is a way to make pre­dic­tions based on in­te­grat­ing ex­ist­ing data, and then sta­tis­ti­cally de­riv­ing a like­li­hood that the pre­dic­tions may be cor­rect.” But, he em­pha­sises, the pre­dic­tions are then to be tested, “and the sys­tems ap­proach val­i­dated.”

Like all pre­dic­tive stud­ies based on com­puter mod­el­ing, the va­lid­ity of the re­sults de­pends on the in­tegrity of the data and the al­go­rithm. If the data are cher­ryp­icked to sup­port the mod­eler’s de­sired con­clu­sions, or if the al­go­rithm is flawed, the re­sults will be in­ac­cu­rate. But it is un­clear from Ayyadu­rai’s ar­ti­cle which data were used, and there is no val­i­da­tion of the model.

Folta of­fers a bril­liant send-up of Ayyadu­rai’s work. “If you de­vel­oped a com­puter pro­gramme that in­te­grated In­ter­net data to pre­dict the lo­ca­tion of Mu­nich, and the pro­gramme told you it was squarely in the Gulf of Mexico, right off Florida, it does not mean that Mu­nich is in the Gulf of Mexico, right off of Florida.” In­stead, it means that you have made a mis­take, in your pro­gramme, as­sump­tions,

the or in­put data – all of which are testable.

To de­cide not to chal­lenge those data, Folta con­tin­ues, and in­stead to “pub­lish a map show­ing that Mu­nich is squarely in the Gulf of Mexico, op­pos­ing all other data and the claims of mil­lions of rather dry Ger­mans, does not mean that you are bril­liant. It means you have ab­so­lutely no clue, or more likely, have some rea­son you want a ma­jor Ger­man me­trop­o­lis to be a two-hour boat ride from Tampa.”

In the spirit of sci­en­tific co­op­er­a­tion, Folta of­fered to col­lab­o­rate with Ayyadu­rai on univer­sity-based test­ing of ge­net­i­cally en­gi­neered corn and soy sam­ples (along with ap­pro­pri­ate con­trols), with anal­y­sis by an in­de­pen­dent lab. Ayyadu­rai de­clined, so Folta will pro­ceed him­self.

The ex­per­i­men­tal data is forth­com­ing. In the mean­time, if you get a han­ker­ing for sauer­braten and spaet­zle, head for Cen­tral Europe, not the Gulf of Mexico.

Newspapers in English

Newspapers from Cyprus

© PressReader. All rights reserved.