Science fraud: from patchwork mouse to patchwork data

1 TheFASEBJ ournal EditorialScience fraud : from patchwork mouse to patchwork dataSummerlin pulled two white mice from the container. Whilethey wriggled and squeaked in protest, he inspected the sites ofthe black skin grafts. Impulsively, Summerlin took his felt-tipped pen out of the breast pocket of his white coat and appliedit briefly to the grafted patches on the two white animals. Theink made them look darker. Then he replaced the mice in thebin and strode out ..FromThe patchwork mouse , an account of William 's 1974 false claim of skin transplantationwithout immunosuppression (1). fraud was so much simpler a generation ago. Allone had to do was to take a felt-tipped pen and colora square patch of mouse skin. The incriminatingpatchwork was also easy to detect. Summerlin's faked"transplants" were discovered by a laboratory assis-tant who washed off the black ink with a ball ofcotton soaked in a little alcohol.

2 Yet the scandal andits upshot in 1974 were just as great as those arousedby the more complex frauds of today. The patchworkincident was described by Jane Brody inThe New YorkTimesas "a medical Watergate" that reflected "dan-gerous trends in current efforts to gain scientificacclaim and funds for research." Indeed, Robert , Summerlin's chief and co-author, was accusedof "manipulating national attention and attracting anenormous amount of money for the institute." (2)Soon afterward, Good stepped down as director ofSloan-Kettering, and that was the end of transplanta-tion without immunosuppression. It sank without atrace in the literature: Summerlin's fraudulent pa-pers have been cited only 3 times in the past 25 years;each a refutation (3, 4).Korean stem cellsNowadays it takes more than a lab assistant with acotton swab to detect fraud .

3 In 2005, Seoul NationalUniversity appointed a broad committee to investi-gate Professor Hwang Woo-Suk's claim that he hadestablished 11 human embryonic stem cell lines bytransfer of somatic cell nuclei (5). The team reportedthat "The data in the 2005 article including testresults from DNA fingerprinting, photographs ofteratoma, embryoid bodies, MHC-HLA isotype matches,and karyotyping have all been fabricated .. the re-search team of Professor Hwang does not possesspatient-specific stem cell lines or any scientific bases forclaiming having created one." (6) To reach this con-clusion, the committee collected cytogenetic samples,checked hundreds of DNA fingerprint samples, exam-ined mitochondrial DNAs and compared dozens ofpolymorphic loci in the genes of alleged donors. LikeSummerlin's patchwork fraud , Dr.

4 Hwang's fabricationwas attributed to scientific pride and overreach. Rich-ard Doerflinger, of the United States Conference ofCatholic Bishops, toldThe New York Timesthat "Hypeand ambition have gotten ahead of the Science ." (7)Soon afterward, Dr. Hwang stepped down from direc-torship of his unit and that was the end of patient-specific stem cell lines in for fraudThe worldwide response to Dr. Hwang's fraud was aspredictable as that to the patchwork mouse a genera-tion ago: pious hand wringing and angry finger point-ing. Now, as then, critics of Science blamed the messen- patchwork Quilt(Courtesy of Michele Fox )5870892-6638/06/0020-0587 FASEBger (the journal) rather than the message (thefabrication). Bold changes were demanded in the wayscience gets written, reviewed, and published; andscientific editors scrambled to get out of the line of one might say that Science journals already have aSarbanes-Oxley code in place.

5 Most journals alreadyinsist that all co-authors sign off on the final manu-script, some oblige each author to spell out his exactrole, and others require each author to affirm theaccuracy of every stage of the manuscript. Nevertheless,the editor ofCellsuggests that it is time for the NationalAcademy of sciences to set strict new standards thatapply to all (8). My colleague, Donald Kennedy ofScience,observed that: "A journal cannot go into au-thors' laboratories in search of fraud ," but hedged hisbet: "More actively, we are committed to examining ourprocesses and ourselves in an effort to extract lessonsfor the future (9)." TheJournal of Cell Biology, whichalready has an elaborate digital process in operation todetecting image- fraud , now plans to erect a screen ofalgorithms designed to spot specific types of imagemanipulation.

6 The software was developed under agrant from the FBI (9). patchwork dataTo determine whetherThe FASEB Journalshould turn toimage scanning, algorithms for detecting statisticalswindles, etc., I've looked at recent cases of fraud andcome up with some tentative conclusions:1)Some scientists will cheat and we'll probably neverknow how often this happens. Young or old, MD orPhD, average or distinguished, male or female, black orwhite or khaki, some scientists will try to pull wool overthe eyes of their colleagues, their reviewers, and )The culture of Science is based on trust, notsuspicion. Great scientists have had fraud committedunder their noses, and good editors have publishedfabrication. Reviewers and editors must have a keennose for swindle, but cannot engage in criminal inves-tigation.

7 As in political life, where I tend to side withliberty over security, in Science , I'd go for trust oversuspicion every )Therefore, we ought not to rely on machines oralgorithms of image or text analysis, but rely on thejudgment of our editors, our editorial boards, and ourreviewers as to whether a manuscript looks as if the datahad been )No measure of "quality assurance", no affidavit ofauthorship, no oath of responsibility or percent effortcan stop a soul hell-bent on self-destruction. And fraudin Science , if not in politics, is always the name of the game is confirmation, Science isself-cleansing: flawed work is soon forgotten and re-mains conclusions are based in part on three recentillustrations of the swindler's of the mouthA study published by the Norwegian oncologist, JonSudb , inThe Lancetof October 15, 2005, concludedthat long-term use of nonsteroidal antiinflammatorydrugs could reduce the risk of oral cancer while expos-ing patients to higher risks of death from heart disease(10).

8 Sudb 's fraud was exposed when Camilla Stolten-berg of the Norwegian Institute of Public Health, whohad access to the primary data, discovered that of the908 people in the study, 250 shared the same birthday!After Sudb confessed to his action, an embarrassedspokesman for the hospital admitted that the data were"totally false, actually totally fabricated. His databasehad been completely fabricated on his computer" (11).The fraud led editors atThe New England Journal ofMedicineto look into two earlier papers by Sudb andsure enough, the journal issued an "expression ofconcern" because two photomicrographs in a 2001paper which purported to represent two different pa-tients and stages of disease were in fact differentmagnifications of the same photomicrograph. Sudb had patched histograms of ploidy to overlay his pho-tomicrographs so that the ruse was not immediatelyevident (12, 13).

9 Keeping withThe Lancet'santifraudpolicy, Sudb 's paper not only affirmed that "JS, JJL,SML, and AJD (Sudb and collaborators) contributedequally to this paper" but also detailed the part eachauthor played in statistical analysis, cutting the sections,writing the paper, etc. Ironically, as required, the paperaffirmed that "All authors approved the final report."Conclusion: So much for quality assurance as to image fraud , Michael Rossner ofThe Journalof Cell Biology, while arguing that "The goal of ajournal editor should be to catch these things beforepublication if at all possible," admitted that hismethods are not yet up to catching tricks of magni-fication like Sudb 's (13).Immunologic toleranceLuk van Parijs, a 35 year old "rising star" in the field ofRNA interference (RNAi), was dismissed from MIT inNovember of 2005.

10 After the whistle was blown byco-workers in his lab, Van Parijs confessed that he hadfabricated data in grant applications, published papers,and in submitted manuscripts. He'd been hired by MITon the basis of strong publications as a graduatestudent at the Harvard Medical School and a produc-tive postdoc stint with David Baltimore at Cal much of his work is under scrutiny. David Balti-more, Cal Tech president, and no stranger to suchmatters, wiredSciencemagazine: "I thought Luk was anexcellent scientist and truly cannot understand why hewould fake anything" (14).The New Scientistengagedexperts to scrutinize two of van Parijs's papers, onefrom Harvard, the other from Cal Tech. The issueremains contested by van Parijs (15). A 1998 paper inImmunitydescribes howFas/FasLandBcl-2,factors588 Vol.