Wednesday, April 2, 2014

Fukuyama, Round 2

Update, 4/3/14 - On a hunch, I looked through major publishers' databases for more recent Fukuyama corrections sharing a specific author's name. Found this (Chartelline C) at ACIEE, alleging "[im]properly processed NMR spectra." Since this author published >40 papers with Fukuyama, more may be coming.


ACIEE, 2014 (above) and 2012 (below)
An astute commenter has clued me in to the latest round of corrections coming from the Fukuyama research group (see 1, 2, 3, 4). Targets affected include lyconadins A-C, ecteinascidin 743, and mersicarpine, and publication dates range from 2010-2013. This time, they're in JACS which, to my knowledge, does not (yet) employ a full-time data analyst like Organic Letters does.

The corrections read much the same as the last raft, released two months prior in Org. Lett., alleging spectral manipulation for multiple intermediates and final molecules, including removal of solvent peaks. Here's one example showing 13C spectra* for intermediate 7, from the synthesis of lyconadin A:

Original spectrum (2011)*

Corrected Spectrum (2014)*
Clearly, there's a bunch of "stuff" in 7 that had to be digitally removed by one of the authors. And this is just one compound; each paper contains several such examples of "improperly manipulated" spectra (direct quote!).

So, what's going on here? I have a theory, although it's a bit of hearsay since I'm not privy to the inner workings of either journal. However, the astute reader will note one author whose name appears on nine of the ten corrected papers from Feb-April...uh-oh.

Was there a bad apple in the Fukuyama lab? More details as I have them.

*Apologies for the different shading, due to my computer and not their SI files.

11 comments:

  1. In the ACIE example, the main 'impurity' removed appears to ethyl acetate. Surely it's more work to doctor the spectrum than just to put the sample back on the vac line?

    ReplyDelete
    Replies
    1. Totally is. Question is, it's not even a lot, why not just publish as-is???

      Delete
    2. It's quite common for referees to demand completely clean spectra, even when it's just a bit of solvent.

      I'm always torn when I see reports like that - it's good that they're looking closely at the spectra, but when the impurities are insignificant, little is gained.

      Could also be that the PI has the attitude of 'give me clean spectra, I don't care how you get them.'

      Delete
    3. No good reason for this, and I know nothing about this specific instance, but pressure can be a bitch. Pressure from the PI and from the group. I'm sure we've all seen and heard about groups that engaged in this behavior, without the PIs knowledge I'd like to believe.

      As a PD a copy of a dissertation from a top 5 synthetic group made the rounds, with clear examples of NMR spectra being doctored. The other "reason" for this, primarily in total synth, is NMR spectra show solvents, then they report 99% yield. There was a backlash about this practice, so instead of drying the sample and accepting a yield of 85% and a deflated ego, its easier to just electronically remove it. Again, pressure to "be the best".

      Delete
  2. What's odd with the Chartelline case is that there were already several other spectra in the SI that had still ethyl acetate in the spectra. It looks like they only removed it from a few spectra for no good reason.

    ReplyDelete
  3. More than 40 publications? As a postdoc, that'd be obscene, but even as a graduate student (particularly in a synthetic lab), that seems really high. One of the grad students in my lab had a lot, but he was not doing synthesis, and publications (or results worth publishing) were easier to get in his research area. Baran, I thought, also had a bunch, so that many pubs is a (theoretically) plausible output in synthetic organic chemistry. Since such output is so rare, perhaps someone might have had (or should have had) an inkling that there was some funny business going on.

    ReplyDelete
    Replies
    1. Hap: An interesting wrinkle in the Japanese system. This researcher did his Ph.D. under Fukuyama, worked 2 years in industry, then came back as a super-postdoc / Asst Prof for nearly a decade. Thus, 47 papers = 8 as a grad, 39 as a junior prof.

      Delete
    2. OK, that makes sense - for something like 14 years, 47 papers is probably a lot (for total syn) but not insane. Sorry.

      Delete
  4. Unstable IsotopeApril 03, 2014 9:51 PM

    I have sympathy for not being able to get completely clean spectra but this seems lazy. That's a lot of impurity

    ReplyDelete
  5. molecular architectApril 04, 2014 8:14 AM

    I agree with Unstable Isotope; sometimes it's extremely difficult to remove all traces of solvent. However, the original spectra here just look bad and suggest the person was too lazy or under too much pressure to go back and get a cleaner / better spectrum. Perhaps it's the latter and I don't think that kind of pressure is what chemists need. I want to see real data, real spectra, real experimental details! What's the point in publishing touched-up spectra? It's akin to the glamour industry in redefining what is normal and requiring everyone to meet the same hyper-idealist standards in the end.

    ReplyDelete
  6. this is clearly a teachable moment, and the researcher should have known better, but i'd let him off with a slap on the wrist myself. removing solvent peaks isn't in the same league as faking data. as for % yield calculations, this is hardly the only way to allow yourself to inadvertently fudge the numbers

    ReplyDelete