MIT professor Ram Sasisekharan made his name on the idea that algorithms and computer models could lead to better and more potent therapies, a promise that launched three biotech companies and attracted hundreds of millions of dollars.

But two treatments purportedly discovered with Sasisekharan’s computational approach are almost identical to compounds that had previously been described by other labs, according to a new paper by outside researchers. The finding casts serious doubts on the integrity of Sasisekharan’s research, the authors claim.

“We looked at exactly two cases, and in both did we find irregularities,” said Tillman Gerngross, CEO of the private biotech firm Adimab and a co-author of the paper. “To me, if you’re sitting in the kitchen and two fat cockroaches walk across the floor, what’s the chance that there’s only two?”

advertisement

In a paper published Monday in the journal mAbs, Gerngross and his Adimab colleagues said two antibody therapies described by Sasisekharan’s lab — one for influenza and one for Zika virus — “show striking similarities” to the past work of other scientists.

Sasisekharan’s influenza therapy, described in 2015 in the Proceedings of the National Academy of Sciences, is nearly identical to one published four years earlier in Science, according to Adimab. The researchers said the Zika antibody, which Sasisekharan’s lab described in Cell last year, bears a strong resemblance to one described in Nature in 2016.

In an email, Sasisekharan said the paper was “inaccurate and slanderous.” There are “fundamental differences” between the Zika antibody he discovered and the one published in 2016, he said. And the influenza therapy in question was designed by Visterra, a biotech company he co-founded, not by Sasisekharan himself, he said.

In a statement, an MIT spokeswoman said that “while federal regulations and MIT policy do not allow us to comment on any particular matter, research integrity at MIT is paramount. MIT has policies and confidential processes in place to assess concerns that might be raised.”

In each example examined by Adimab, Sasisekharan’s lab did not publish the amino acid sequences of the antibodies it purportedly designed. In order to check the work, the Adimab researchers sought out the resulting patent applications and cross-checked the details on GenBank, an open-access database of sequences. That’s where they found similarities to earlier work, according to Gerngross, who is also a professor of bioengineering at Dartmouth College.

“We find it difficult to view these authors’ approach in any light other than an intent to mislead as to the level of originality and significance of the published work,” the Adimab group wrote.

William Schief, a professor of immunology at the Scripps Research Institute who reviewed the Adimab work before it was published, said the paper makes a “very strong case.”

“If you look at the original [MIT] papers that reported these antibodies, they don’t give a really clear description of how they identified the epitope or how they designed the antibodies,” he said.

The implications of Adimab’s paper stretch beyond academia. Visterra was developing the flu antibody when it was acquired by the Japanese drug maker Otsuka for $430 million last year.

A spokesman for Otsuka did not immediately respond to a request for comment Tuesday.

Sasisekharan, who is a member of MIT’s prestigious Koch Institute for Integrative Cancer Research, has received prominent recognition for his work, including from the National Institutes of Health.

Leave a Comment

Please enter your name.
Please enter a comment.

  • Wow, I have never seen a story like this before?
    “Connecting the sequence dots: shedding light on the genesis of antibodies reported to be designed in silico”
    ” Including conservative substitutions (“positives” as in the default settings for BLAST22″
    “….. reveals an even closer relationship with a similarity of 96% and 88%…..”
    Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ. Basic local alignment search tool. J Mol Biol. 1990;215:403–10. doi:10.1016/S0022-2836(05)80360-2.
    [Crossref], [PubMed], [Web of Science ®], , [Google Scholar]
    “….. reveals an even closer relationship with a similarity of 96% and 88%…..”

    • You shouldn’t trust them – you should read the paper, check the facts and draw your own conclusions
      I like the notion of :
      Materials and Methods will be supplied upon request

  • Few months back one guy in Cambridge Uni forged data and published in some top journal Nature/Science/Cell and that postdoc’s boss Prof totally trusted his work and didn’t crosscheck vigorously.

  • If this is true, this person is an absolute embarrassment to the scientific community.. and should be excommunicated immediately -.-“

  • As I use to say,
    cheating is the most effective way to make a career in science ….
    or to become president!

A roundup of STAT’s top stories of the day in science and medicine

Privacy Policy