Skip to main content

Life As We (Don't) Know It

Arsenic Bacteria Ride Again. (Or Don’t).

You may not have heard much about the arsenic-bacteria controversy recently, but you’re about to hear quite a bit more. Rosie Redfield of UBC, one of the fastest and most vocal critics of the original paper, has been trying to reproduce it in her own group. There’s a manuscript in preparation, but since she’s been blogging on some of the progress, the import is clear: it hasn’t been going well for the “bacteria can take up arsenic in their biomolecules” hypothesis. Scrolling back at that link will give you the story.
Here’s a summary at Nature News (with a clarification from Redfield on one point). I look forward to seeing how this plays out – but remember, the startling results always have to prove themselves by happening again. Einmal ist keinmal.
Update: there’s another story here, too. Redfield has been posting results as they come along, in a very prominent example of “open science”. The first question is: will this affect journal publication? That is, will some editors look askance? The second point is to be found in that Nature News article, where Felisa Wolfe-Simon refers to those “website experiments”, and how she basically can’t discuss them until she sees them in a journal. Note that it’s not “the UBC experiments” or “Redfield’s experiments” – they’re “website experiments”, and thus (apparently) have more to prove.

19 comments on “Arsenic Bacteria Ride Again. (Or Don’t).”

  1. RB Woodweird says:

    “We are thrilled that our results are stimulating more experiments from the community as well as ourselves,” first author Felisa Wolfe-Simon, now at the Lawrence Berkeley National Laboratory in California, wrote in an e-mail to Nature. “We do not fully understand the key details of the website experiments and conditions. So we hope to see this work published in a peer-reviewed journal, as this is how science best proceeds.”
    Interesting that Felisa pulls out the peer-reviewed or bust card, seeing as how her results were communicated via press conference.

  2. Carmen Drahl also had a great piece about this over at C&E. The sad thing as someone noted there was that the arsenic proponents will fight a rearguard battle and gradually retreat into the shadows rather than capitulate outright and admit their fault; the whole thing is in danger of being slowly forgotten rather than going down in the history books as a glaring cautionary tale (which is should).

  3. Derek Lowe says:

    RB, that’s interesting. In between posting this on the train ride in to work and getting here to my office, I had exactly the same sort of thought, and thus the updated post. Then I read your comment!

  4. SP says:

    There are some large subsets of journals (maybe the ACS family?) that require you to certify that the results have not been previously published in a public forum.

  5. chirality says:

    The problem with this whole arsenic DNA debacle is that Felisa Wolfe-Simon’s claims are not falsifiable. Therefore, this is not science. Initially, FWS led everybody to believe that arsenic was replacing phosphorous in DNA based on very flimsy evidence and wishful thinking. Faced with the well-deserved skepticism, FWS has changed her tune. Now the claim is that some biomolecules produced by her namesake bacteria incorporate As instead of P to SOME DEGREE. Therefore, to disprove this claim, one would have to show that no biomolecule incorporates any (we are talking single atoms here) As instead of P. Good luck with it! This is very much like the hexacyclinol parody. Everybody knows it is rubbish, but it is immensely difficult to prove it.
    Personally, I would like to see this arsenic thing retracted just to deflate FWS’s ego – who on Earth names an organism after oneself?

  6. David Sanders says:

    The text of the original Science article misrepresents the data (Supplementary Table 1, for example) as does the authors’ Response to the Technical Comments. The authors have committed misconduct both in the original publication and in their persistence in their pseudoscience after it has been demonstrated incontrovertibly that they provide no evidence for their claims.

  7. cirby says:

    I figured it out.
    They’re not arsenic-based organisms.
    One of the bacteria just took out an insurance policy on the rest, and is trying to collect on it by poisoning them.

  8. anchor says:

    #5, well said! How can one prove the negative?

  9. paperclip says:

    As to the the first question in Update, Dr. Redfield says that she has been blogging about her work for about 6 years, and it hasn’t affected her ability to publish. (But her arsenic work has surely received much more attention than her other projects, so I’m not sure if past experience will be a reliable predictor.)
    There’s been a lot of talk about “open science” lately. The New York times covered it recently, and I noticed that most of the commenters were critical, many of them envisioning the process as a Facebook-style popularity contest. Whatever the case, for most of us article-writers, and I mean 99+%, things will remain the same. There are far too many articles for each of them to get attention from blogs and the like.

  10. spoons says:

    @10 They say Facebook style popularity contest as a bad thing? I could easily imagine having a organized social network of the scientific community with their real names attached to all their work and comments, with information coming out in a relative real time.
    Sure that is a bit of a pie-in-the-sky scenario I cooked up but some peoples apocalyptic visions of Facebook are a bit overblown in my opinion.

  11. WesvewatwolWeceptow says:

    @Derek: u take the train? now that pfizer’s in town why not hop on the chopper?

  12. Hap says:

    I hope Dr. Wolfe-Simon doesn’t become a prosecutor.
    “Your Honor, the defendant haven’t proven that he didn’t do it, so he should be found guilty. In addition, the defense insists on bringing up inconsistencies in our case in public and not behind closed doors as they should.”
    (mumbled conversation with the judge)
    “What? You mean we have to, like, prove our case, or at least not posit a story with more holes in it than a pound of Swiss cheese? That’s preposterous.”

  13. Anonymous says:

    @5 chirality: I think the same type of person who names an institute after oneself.
    In the same vein, @13 Hap: I agree, but can we leave out the sexist undertones? (If FWS had been male, would you have written the “like”? I hate when people conflate her femaleness with her ignorance.)

  14. Falanx says:

    To #14:
    Really? I saw no conflation. I saw the stylised representation of the words of a moron. I’d have phrased it equally so if ‘Felix Wolfe-Simon’were our subject.
    What I did see was an unfounded assumption, though…

  15. Dave says:

    The Wolfe-Simon results were communicated by press conference, but only after they had been peer reviewed and published. Redfield’s have not been. What is so hard to understand about that?

  16. Hap says:

    Sorry – the “like” was too feminine-implying. I would have had the same contempt for whoever made the argument (and since it’s been a staple of ID, I would have had opportunity to use it against those of the other gender), but I didn’t intend the sexual slight.
    I used to use “like” too often, though that was in sixth grade and twenty-five years ago.

  17. Hap says:

    17: Yeah, but so were cold fusion, NaH oxidation, and the JJLC hexacyclinol “synthesis” (and at least the last two had more than enough problems that their publication should have been aborted well before we ever saw them). The “reactome” paper was also published in the same journal, with chemistry in the schemes that would get you failed in a first-semester orgo or biochem course and a theory of biological action that seemed…sketchy. (See this comment and this comment for reasons why.)
    Peer review is supposed to keep out crap, but it doesn’t always do even that, even at the upper echelons of journaldom. It shouldn’t be seen as a talisman to reject arguments of incorrectness. It lowers the likelihood of a paper being complete crap, but doesn’t eliminate it, let alone validate the content (since you may not find fraud by peer review). There is also the step of having others reproduce your work. (You could argue that with other authors, the reproducibility of work is taken more for granted than in this case, and that that indicates a bias of some sort. People could be biased, either by sex or by youth, although the facility of exchanging ideas and results in some form would seem to be a greater factor in the assumption of fallibility.)
    Even if a paper is peer-reviewed, and reasonably so, it’s still possible to claim lots of things in a press conference that the data in the paper can’t support. There are good reasons that publication by press release is so poorly regarded – you’re not supposed to make claims that your data can’t cash.

Comments are closed.