press conference

Mike Rossner has built a career addressing problems related to image data integrity, such as the 2006 Korean stem cell fraud case pictured here being discussed at a press conference.


Bringing image manipulation to light

In a sense, one question sealed the deal for Mike Rossner’s entry into scientific publishing. In the waning days of his biology postdoc in Melbourne, Australia, as he considered leaving research and wondered what he might do instead, he took a frenetic 10-day trip spanning London and San Francisco, meeting with journal editors to see if publishing might be his thing. In London, an editor asked him, “Are you the sort of person who will go up to someone you’ve never met before and introduce yourself?” Without missing a beat, Rossner answered, “Yeah, no problem,” he recalls. “Well then, you might be right for publishing,” she said. She then handed him the phone number of her daughter, who, along with her husband, was a recent transplant to Melbourne. “Call my daughter; she doesn’t know many people,” the editor instructed. Determined to prove that he had the people skills required for a publishing career, and to help out someone new to town, Rossner complied. 

It’s this mix of initiative, mentorship, and serendipity that has propelled Rossner’s career forward over the past 20 years, many of them spent at Rockefeller University Press (RUP) in New York City. After the journals there restructured in 2013 and Rossner lost his job, he embarked on a new, more risky venture: his own consulting company, which advises journals and institutions about how to handle image manipulation in published papers and how to screen manuscripts before publication. It’s a corner of the publishing world Rossner came to know unexpectedly well, and it is where he has found a double dose of satisfaction in helping to maintain the integrity of published research.

Carving out an alternative niche

Early in his training, if he gave it any serious thought, Rossner leaned toward a career in scientific research. As an undergraduate at Princeton University in the mid-1980s, he joined the second class of majors in a brand-new program: molecular biology. “There was certainly a buzz” about the field, he says, and he was entranced. In graduate school at the University of Edinburgh in the United Kingdom, he landed in the lab of Sir Kenneth Murray (then without the knighthood), a pioneer in hepatitis B research and one of the developers of a vaccine against it.

Murray—and Rossner, it turned out—had other interests too. The professor “took it upon himself to teach each of his graduate students how to write,” Rossner remembers. “That was extremely important to him.” Rossner’s first paper cycled through dozens of drafts with the older scientist. “It wasn’t just every sentence; it was every word,” and as each draft sailed back with more red markings, Rossner tried to squelch his frustration. “You had to express your thoughts clearly in as few words as possible. … At the time,” he says, “I didn’t realize what was happening to me, that I was being mentored in how to communicate science.”

Mike Rossner

Mike Rossner

Mike Rossner

Rossner held his adviser in awe and treasured those evenings when Murray would emerge at 9 or 10 o’clock at night from his office, a strong cup of coffee in hand, to chat with his graduate student. During one such conversation, Murray suggested that it was time for Rossner to start thinking about a postdoc. He urged Rossner to consider a position in a Melbourne lab where one of his own former postdoctoral fellows was working. Rossner, presuming he’d stick with bench research for the foreseeable future, quickly agreed, even though the lab in question focused on cancer biology, far afield from what he’d been studying. Rossner’s attention hewed not to unraveling specific scientific mysteries, but rather to new technologies and people who might guide him. “[I] hadn’t chosen cancer biology for the topic; I chose it for the mentor,” he says. Now, he reflects from his home office in a leafy town outside San Francisco, California, he remembers that even during his training, “there was never a topic, a biological question, that I was passionate enough about that I was going to write the grants and hire the people and run the lab.”

Still, a publishing career didn’t come into focus until his postdoctoral fellowship at Melbourne’s Walter and Eliza Hall Institute of Medical Research. The institute brimmed with a mix of nationalities and, by extension, varying expertise in English. Over 3-and-a-half years there, Rossner helped many polish their manuscripts before they were submitted, a task he found he enjoyed. “You had this nice, immediate reward,” he says, in contrast to the sometimes glacial pace of lab research. Word traveled that he had editing skills to offer. “[I] had people say to me, ‘You should do this for a living.’”

His direct supervisor wasn’t a fan of nontraditional careers, but the head of the institute, Gustav Nossal, proved more approachable. Rossner stopped by Nossal’s office one day, and the older scientist embraced the younger one’s interest in scientific publishing. “To get that response from the director of the institute, 25 years ago, was extremely encouraging,” Rossner remembers. Nossal wrote letters of introduction to various journal editors he knew. Rossner followed up with phone calls—often at 3 or 4 o’clock in the morning in Melbourne to catch those in Europe and North America during daylight hours. 

Then he embarked on that 10-day trip around the world, sitting down with editors whether or not they had jobs to fill. A meeting in San Francisco with the editor-in-chief of Science yielded pleasantries but no offer. In New York, the editor-in-chief of Cell handed him a recent issue of the journal and asked whether the papers in it deserved to have been published there. “I had to question his judgment in front of his face,” says Rossner, who still squirms mentally when remembering the conversation. But “if somebody’s thinking about publishing, they need to be prepared to make those sorts of judgments, even in an interview.”

Next, Rossner flew to London for an informational interview with the head of Current Biology. That editor mentioned that Rossner ought to chat with the person in the office next door, a former editor at Nature, because her interest in molecular biology aligned with his own. She was the person whose daughter had moved to Melbourne—and who helped get him his first scientific publishing job. That break came when a position opened up at Chemistry and Biology, at the journal’s tiny San Francisco outpost. The London editor put in a positive word for him, and he got the job. “Keep reaching out to journal editors,” he advises eager editors-to-be, “because [they] will know of positions opening up, even at other journals.” 

Rossner spent the next 3 years learning the ropes, which involved much more than fielding manuscripts from the office. He traveled to conferences, visited labs, and networked endlessly. “Everyone said, ‘Oh, do you miss the lab?’” Rossner remembers. The answer was no. Publishing good research was reward enough, even though it wasn’t his name on the papers.   

Three years later, in 1997, he moved to RUP to become managing editor of the Journal of Cell Biology (JCB). It was a time of major transition for JCB and hundreds of other journals: They were shifting from paper to electronic submissions, and within 5 years the entire workflow at JCB was electronic. It was this change, and the headaches that accompanied it, that opened up a new career direction for Rossner.

A change of plans

In 2002, as JCB was cementing the move to electronic publishing, its editors often struggled to manage files that arrived in “weird formats,” Rossner says, an issue especially common with image files. “We used to go back to the authors and say, ‘Your paper’s accepted, but we’ve got to get this in a format we can use before publishing it.’” One day, JCB was on the cusp of publishing a paper when Rossner realized the problem had flared again and the figures—Western blots used to detect specific proteins—were in an unusable format. But this particular paper didn’t have many of them, so rather than go through the trouble of contacting the authors and waiting several days for new files, Rossner decided to take care of the conversion himself. 

“I opened up one of the figures, and there was an obvious black box around one of the bands” in a Western blot, he says. Rossner realized that this particular band “had been adjusted.” At his request, the authors sent in their original files, which didn’t match what had been submitted. The paper’s acceptance was rescinded. 

But Rossner wondered whether he had a much bigger problem on his hands. He and his colleagues began combing through other submitted manuscripts and “started seeing problems all over the place.” Within weeks, JCB became the first journal to announce that it would begin screening for image manipulation in all accepted papers. At JCB, about 1% of accepted papers had manipulated images that affected their conclusions; another 25% had some sort of manipulation that violated guidelines. These numbers held steady during his time at RUP.

“Images are data, I think that is key,” says Rossner, who believes the rules are pretty straightforward. “You want your data to represent, as accurately as possible, what you saw.”

Catching manipulated images became a passion, even though finding them inevitably leaves him with a pit in his stomach—another research group that needs to be confronted, another paper whose acceptance will be rescinded. After the landmark fraud case involving manipulated images and stem cell scientist Woo Suk Hwang broke at Science, journal editors began traveling to Rossner's New York office to learn how to study images and detect problems, and he often spoke out and wrote about the issue.

Then 3 years ago came a detour in what had been a steady career trajectory: Rossner’s position at RUP was eliminated, and he was out of a job. But his interest in image manipulation and software related to it proved a boon. While at RUP, he had worked with a company named Glencoe Software, which creates tools to view, analyze, and manage scientific data. When he lost his job, they offered him a position. 

After a couple years there, he decided to try his hand at something new: his own company, Image Data Integrity (IDI), which he launched in 2015. IDI melds Rossner’s experience in biology, publishing, and software to address two needs: helping institutions in their internal investigations by offering an independent opinion on suspected image manipulation, and assisting journals who may not have their own in-house detectives or who want to establish a screening process. The plunge has been a little scary, but he’s happy with his decision—even though he spends about half of his time drumming up new business.

For those curious about a career in scientific publishing, Rossner says that although the world of journals has changed since his early days—with an explosion of new titles and a trend toward open access, which he embraces—the basic principles haven’t. Communication matters—at his first publishing job, he says, his boss nurtured Rossner’s communication skills and diplomacy in sensitive situations, literally looking over his shoulder “at every email I was sending out to a member of the community.” Aspiring editors must also learn to be confident about making rapid judgments. Some newly minted editors take hours to pore over a submission, which isn’t practical when another 50 are flooding in during that time. 

Finally, he notes, “there is a great reward to putting out a substantial piece of work every month,” or however often a journal publishes, but “it’s not your work. … The primary research is never yours.” The editors operate quietly behind the scenes, wielding influence but rarely basking in the glory of science. “Your name is not on the paper.” It’s a tradeoff he’s been happy to make.

Follow Science Careers

Search Jobs

Enter keywords, locations or job types to start searching for your new science career.

Top articles in Careers