Coffee Break: Fossil Fuels and Health, EPA, AI that Works, and a Belated Retraction

Part the First: The Healthcare Costs of Fossil Fuels.  One would think this is a no-brainer.  Nevertheless, the Current Administration has its stone cold heart set on gutting the Office of Research and Development in the Environmental Protection Agency.  I suppose it would be churlish once again to note the EPA is one of the signal achievements of the last liberal Administration in the US – that of Richard Nixon.

I have not had the chance to discuss this with my oldest friends, who are now mostly MAGA Men and MAGA Women.  We grew up in the 1960s, when our air was unfit to breathe whether we could smell it or not.  At one time the Spanish moss disappeared from our live oak trees because of the foul air.  When the local industries were required to curb their releases because of the Clean Air Act (Johnson, 1963; Nixon, 1970), the moss came back, after 1970 Amendments:

Required EPA to determine which air pollutants posed the greatest threat to public health and welfare and promulgate National Ambient Air Quality Standards (NAAQS) and air quality criteria for them.  The health-based standards were called “primary” NAAQS, while standards set to protect public welfare other than health (e.g., agricultural values) were called “secondary” NAAQS.

The fish came back to our tidal rivers as a consequence of the Clean Water Act (Nixon, 1972).  The healthcare costs of polluted water require little explication.  The healthcare costs of bad air are often ignored, but as this article at SBM notes:

Burning fossil fuels releases sulfur dioxide, nitrogen oxides, particulate matter, carbon monoxide, and mercury into the atmosphere.  The cumulative effects of these pollutants exacerbates respiratory illness such as asthma and COPD, increases the risk of cardiovascular disease and cancer.

A recent study confirms that these effects are not just theoretical, but can be directly measured. They looked at the rate of ED and hospital admissions for asthma and other respiratory illness for the three years prior to and after the closing of the Shenango coking plant located on an island in the Ohio River near Pittsburgh, PA. They found that after the closing there was an immediate 20% decrease in weekly respiratory ED visits, and a 41% decrease for pediatric asthma ED visits. This was following by a 4% decrease per month for the duration of the study. So there are both short- and long-term benefits to respiratory health.

At the high end of estimates is a study (pdf) by The Medical Society Consortium on Climate and Health, who estimated the annual cost to the US at $820 billion.  This includes the direct effects of pollution and the resulting climate change.  This does not include non-healthcare related costs.  This does include 107,000 premature deaths each year.  On average this costs each American $2,500 in increased medical bills annually.  While it’s possible to quibble about the methods used and what exactly should be included, the estimate is not unreasonable when trying to account for the total societal cost. It is also likely an underestimate for the factors it’s including due to incomplete reporting.

Of course, coking plants and other polluters such as my first employer during my days as a heavy chemical worker have simply been moved to other countries with lax environmental laws, but the point remains.  The effluents we continue to release into our environment, not the environment, make us sick.  And the point of closing the EPA Office of Research and Development is to make the problem “disappear.”  This is no different from reducing pandemic reporting or to be a bit more prosaic, turning up the car radio so the strange knocking from the engine compartment goes away.  But back to my many MAGA friends from our formative years, have we really come to this?  Rhetorical question.

Part the Second: The Attack on Scientific Research Continues.  The Food and Drug Administration (FDA) will soon be unrecognizable.  The current director, Dr. Marty Makary, late of Johns Hopkins University School of Medicine but still a member of the National Academy of Medicine, has done his research.  Last year he published much of this in his book Blind Spots: When Medicine Gets It Wrong, and What It Means for Our Health.

I read it with the intention of discussing it here, but it was not worth the trouble, especially after doing the same for Casey Means MD, still the Surgeon General nominee last night, and her book Good Energy that has become MAHA holy writ (for the rich and famous and rich and not so famous).  Dr. Makary is not wrong at the most superficial level, but his blind spots have been covered better by others.

SBM is on the case once again.  Dr. Jonathan Howard quotes Dr. Makary:

We have a lot of data and it may not necessarily be the traditional 50 year randomized control trial follow up.  It’s data from families that say their kids have been acting with bad behavior…and they eliminate the petroleum-based food dyes and the behavior improves.  That is data.

Actually, no.  That is “anecdata.”  As an aside, my lesson for my graduate students was always this: Once is an anecdote.  Twice is data.  Three times is a result, but if and only if your subsequent experiments build on this particular result.  In any case, this Dr. Makary’s trope is no different from “post hoc ergo propter hoc” reasoning.  This “thinking” has been responsible for much of the vaccine hesitancy we see today, especially after the pump was primed by the former Dr. Andrew Wakefield with his publication of a spurious link between MMR and autism in one of our leading medical journals, The Lancet.

A bit later:

I personally know of people who have been injured by the vaccine.  I personally know of friends who have lost a loved one from the mRNA COVID vaccine…People have a right to be angry.  They have been deceived … I would ask people to be patient with us as we do this the proper scientific way.

To which I could also retort, “I personally know someone who was the “one-in-a-million” injury caused by the attenuated polio vaccine.  Her “mild” childhood polio left her with a limp and other problems that have made her life difficult at times.  But not once has she ever lamented the polio vaccine.  I asked.  She is a very good young friend who is now an outstanding urologist and soon-to-be a professor in a leading medical school.  She does not consider herself the object lesson for the proscription of the polio vaccine.  And neither to his credit is Senator Mitch McConnell, who was the only Republican Senator to vote against confirmation of the current Secretary of Health and Human Services due to his fulminate vaccine denialism.

Director Makary goes way too far in his ridiculous statements about viral pathophysiology:

While Dr. Makary spreads rumors regarding vaccine-harms, when it comes to viruses, he does the opposite. Instead of telling tall tales, he rejects reality.  Nearly identical examples about measles and COVID are below.  Compare the casual way he dismisses dead children with the seriousness and concern he showed for supposed vaccine injuries. Hoping to numb parents to potentially grave threats, he claims that viruses don’t kill healthy children.  This obviously false, healthy children have died from COVID and measles. Moreover “soft eugenics” is appalling, and viruses can hurt kids without killing them.

Of the millions who have died of viral illness over the course of human history, most of them were healthy before their viral infection – measles in post-European contact America, for example.  But we do not have to go back that far to remember.  One of my best friends, a swimmer and diver and brilliant landscape architect who could have been the model for Adonis, was perfectly healthy, that is until HIV destroyed his immune system and he died at the age of 39 – four years before HAART became available.  I would go further and say the “soft eugenics” of the current madness has hardened considerably and is likely to become much harder before it relents.

Regarding food additives, yes, remove them.  But this will not Make America Healthy Again any more than replacing high-fructose corn syrup with cane sugar will make soft drinks “healthy” (the problem of dietary fructose is complicated, and I hope to get to it one of these days).  Soft drinks will never be anything more than a treat.  The problem is not the sugar, wherever it comes from.  The problem is the serving size.  When I was a kid, that was 6.5 ounces in a returnable bottle that was reused as many as twenty times.  Now the serving size can be a 32-ounce plastic cup that often comes with free refills.

Part the Third: AI Works, When the Training Set Is Valid.  Thirty years ago, when I was the oldest postdoc in the oldest Department of Biochemistry in the United States, the dream of protein structure prediction from the primary sequence of amino acids was just that.  But many of the scientists around me were working hard on the project.  The Wednesday morning Biophysics Seminar was more than an opportunity to eat doughnuts and bagels for free.

This weekly Clash of the Titans was both educational and entertaining.  Much of the discussion came directly from what had happened at the most recent CASP meeting.  CASP is the “Community Wide Experiment on the Critical Assessment of Techniques for Protein Structure Prediction.”  CASP led directly AlphaFold, for which three scientists were awarded the Nobel Prize in 2024.  AlphaFold has made previously unimaginable research possible.  So naturally, the National Institutes of Health plans to discontinue support for CASP.

But CASP was the catalyst that made AlphaFold possible and will be necessary for future developments to remain in the public domain.  So naturally, this is unlikely to matter as the neoliberalization of all science proceeds apace:

For the past several years, CASP has gotten around $639,000 from the National Institute of General Medical Sciences at the NIH, which supports two full-time employees. They coordinate the biennial international conference/competition where computational modelers test their (now largely AI-powered) methods on an even playing field. The grant, administered through the University of California, Davis, is running out with no news of a renewal, and both employees have received termination notices for Aug. 8.

Rather than being part of a targeted defunding, the grant seems to have gotten swept up in the general chaos surrounding biomedical funding in the U.S., according to Moult, a cell biology and molecular genetics professor at the University of Maryland.  He thinks that the organization will eventually get NIH funding again, especially since the experiment’s bent toward ensuring that scientific claims are backed by evidence fits into the current federal administration’s stated priorities.

“The emphasis from the administration is we want ‘gold-standard science,’ which is reproducible, transparent, rigorously [executed], and that’s almost a description of what CASP does in this particular area,” he said.  “Nobody has told us this, but it seems blindingly obvious to me it’s a really good fit for what, at least on paper, the administration wants to achieve.”

If CASP falls apart, it will be hard to get up and running again, he warned.  It would survive, but “deeply wounded in a different kind of setup,” he said, and the U.S. would lose its leadership in this area.  Without U.S. federal funding, a European or Chinese organization would likely pick up the reins in some way.

The locution “gold-standard science” is false.  The “gold standard” is nothing but a fetish, although it does fit in the current White House.  And there is no doubt leadership in this essential discipline would probably pass to China or perhaps a European consortium if public support for CASP ends.  This is no way to run scientific research.  But here we are.

Oh, and back to that Biophysics Seminar.  The doughnuts and bagels were not exactly free.  Regular attendees were required to participate (with personal death being the only excuse not to show up on your scheduled day).  My presentation was not about the technical details of protein structure prediction, which were and will remain beyond my ken.  So, I went back to the original paper in which Christian Anfinsen, a sometime member of my then department before my time, showed that the amino acid of the small enzyme ribonuclease was necessary to make the active protein.  Anfinsen was awarded the Nobel Prize in Chemistry in 1972.

My talk made the case that the amino acid sequence was necessary but not sufficient to specify/determine an active protein structure.  Anfinsen successfully “renatured” ribonuclease only when he included a small amount of a common reducing agent in his preparation.  These conditions allowed the protein to fold and become active, but only because they mimicked the interior of the cell.  I have never had a worse case of stage fright in my life.  The audience of 25 or 30 included three or four members of the National Academy of Sciences and one Nobel Laureate.  They appreciated my perspective but thought I was making too much of the importance of the cell.  Perhaps, but all life is cellular or it is not life.  Alas, In Search of Cell History was still twenty years in the future – highly recommended for all biology nerds in the community, as are any other out-of-print books by Franklin Harold you might find in the used book store.

Part the Fourth: Arsenic Cleanup in Aisle 2010.  Finally!  After nearly fifteen years one of the most remarkable papers to ever have been published in a legitimate scientific journal, in this case Science of all places, has been retracted.

To summarize, a group led by Felisha Wolfe-Simon isolated a bacterium from Mono Lake in California.  They “showed” that this bug could grow without phosphorous by substituting it with arsenic (one element below phosphorous in the periodic table).  There can be few more extraordinary claims about a living organism.  We are all, from bacteria to animals, made of oxygen, carbon, hydrogen, nitrogen, sulfur, phosphorus plus an assortment of trace elements.  Let us start with DNA.  The “sugar-phosphate backbone” of DNA forms the legs of the helical ladder with the bases that make up the information-rich codons on the inside.  The paper implied that the DNA in this bug is made of a “sugar-arsenate backbone.”

This is complete and utter nonsense, not to mention impossible if our understanding of the evolutionary biology of life on Earth is correct (i.e., there is one common universal ancestor of all life).  If this were true, the bacterium would have enzymes that worked on arsenates rather than phosphates in a parallel biological universe.  Extraordinary claims require extraordinary evidence, of which there was none:

After the paper was published, chemists and biologists took aim at it in journal submissions and on social media.  Chemists said that if arsenic were incorporated into DNA’s backbone, the bonds would be so unstable that they would fall apart in water in less than a second.  Microbiologists, including Redfield, pointed out flaws in the work, such as that the bacterium’s growth medium contained enough phosphate contamination that, despite the team’s effort to prove that the organism could live on arsenic, it was probably still phosphate dependent.

“There were very, very powerful reasons to think that the result must be wrong,” Redfield says.

In May 2011, Science published eight technical comments, one of them from Redfield, criticizing the paper, alongside a response from the authors refuting the comments.  The next year, the journal published two studies, including one from Redfield’s laboratory, attempting — and failing — to reproduce the results using bacterial samples from the arsenic-life team.  But Science did not retract the paper.  Until now.

The authors of the paper disagree with the retraction.  Their reasons, perhaps the only reason, the paper has not been retracted until now can be summarized as “there was no fraud in the paper.”  While this may be true, the paper was as mistaken as a biology paper can be.  The data were unconvincing to everyone except the authors and the reviewers.  Regarding the latter, I would imagine they want to remain anonymous.  Regarding the editorial process that allowed this paper to be published in the world’s second-leading scientific journal after Nature, the mind fairly reels.  But the work was supported by NASA, which has been a bureaucracy in search of a budget since Apollo 18 was cancelled.  Stuff happens.

What now, wonders Ivan Oransky of Retraction Watch?  The scientific literature is filled with papers that turned out to be wrong.  Carl Gajdusek was awarded a Nobel Prize for identification of “slow-acting viruses” that cause Kuru (he was also later identified as a pedophile).  It turns out that prions, infectious proteins, are the cause, and for that discovery Stanley Prusiner won a second Nobel Prize for the same thing.  Prusiner’s results were “heretical” but correct.  Wolf Simon’s results were heretical and incorrect.  It is not necessary that Gajdusek’s papers be retracted; prions are real and behave like slow-acting viruses.  Egas Moniz won a Nobel Prize in 1949 for “his discovery of the therapeutic value of leucotomy (prefrontal lobotomy) in certain psychoses.”  His papers do not need to be retracted.  No one cites them in support of the return of the lobotomy, and they provide an example of the dangers of what can only be called stupid science.

A case can be made that the retraction of the Wolfe-Simon paper is superfluous.  But the case is very weak.  Their assumption the bacterium was growing in the absence of phosphate was immediately shown to be wrong.  Remember, bacteria are very good at “making do” under very harsh, starvation conditions.  That the culture medium used was “phosphate-free” was absurd, as any competent biochemist or analytical chemist knew from the start.  Neither were the techniques used to prove the thesis were up to the task.  Isolating DNA from the organism would have been a trivial task.  This was not done but it would have shown that its DNA was no different in content from that of any other living organism.

Not so long ago, a group of structural biologists (x-ray crystallographers) solved several protein structures.  It turned out the structures were mirror images of the correct structures because of an error in the program used to analyze the data (something like a negative sign in place of a positive sign).  It was an embarrassing error.  No fraud was involved but the papers were retracted and the reputations of the authors suffered not.

Still, it is necessary that scientists, reviewers, and editors pay close attention to the business of scientific publication, which in the era of pay-to-publish open-access “journals” has been polluted, perhaps beyond recovery in the near term.  The ultimate problem with such nonsense is that it gives science deniers semi-solid ground to stand on.

Maybe we need a “Clean Journals Act” for the twenty-first century?