Thread by @SciencePunk

Something that has been troubling me for the last year or so: why is the world's most famous geneticist palling around with a sketchy startup hawking pseudoscientific treatments in offshore clinics?

George Church is a professor of genetics at Harvard university, and a scientific advisor to BioViva, a US 'startup' whose CEO famously injected herself with telomerase in 2015 in a bid to reverse her ageing. Note, he is also a shareholder in BioViva.

He defended this position in the past, telling the Guardian in 2016 “I advise people who need advice, and they clearly need advice”, which is honestly a total zinger of a line theguardian.com/science/2016/j…

But Church offers more than advice. Last year, he co-authored a research article with BioViva CEO Parrish, published in PNAS, on a new telomerase-based anti-aging gene therapy, tested in mice:
pnas.org/doi/epdf/10.10…

That article published through a ‘contributed paper’ pipeline that Church can access as a member of the National Academy of Sciences.

It seems unlikely it it would have been published in such a prominent journal otherwise, as BioViva’s other papers are found in Maples Scientific journals, a label which boasts of high acceptance rates, low turnaround times, and is filled with barely coherent nonsense.

(don't believe me? How about Bittmann et al.'s "Zorbing in Impaired Children: An Innovative New Alternative for Better Self-Consciousness") maplespub.com/article/Zorbin…

When the paper article was published in early 2022, it was immediately pointed out that of the two supposedly independent reviewers, Noam Maoz worked for one of the authors, and Bill Andrews was himself part of BioViva. Whoopsie!

PNAS withdrew the paper and it was released again in August, with a correction that claimed Andrews did not become involved in BioViva until after the paper was published, despite the fact that Andrews has been involved with the company since its inception pnas.org/doi/10.1073/pn…

Andrews was the one who prepared Parrish's telomerase therapy in 2015. (Though he later told me he thought she probably took 1000th of an effective dose...)

The correction also takes the opportunity to distance Harvard from the paper, noting that Church’s contributions “were conducted as part of his role as a consultant for BioViva USA, Inc. and not as part of a project administered through Harvard Medical School.”

OK fine, but here's where it gets sketchier

During the course of making our podcast on BioViva, Buying Time, we had a cell biologist take a look at the PNAS paper, and they noticed some... irregularities with the illustrations.

Take this blot for instance. On the left, the follistatin (FST) gene treatment, on the right is the TERT (telomerase) treatment. Notice anything odd? Does the TERT it look.... drawn on to you?

Aside from the lack of feathering, the TERT sample is a different colour. The black on the left is RGB (29,29,29). On the right it is (0,0,0), the value you’d pick if you thought something was black. Ramp up the brightness to see it:

Then there's this image, showing the expression of the treatments in the mice. The TERT sample isn't just a different colour, it's a different encoding. Normal scans in CMYK, yet TERT sample rendered in RGB. On the same scanner??

Unlike RGB, CMYK codes black as a separate channel to the colours, so you can desaturate the image in Photoshop and see how the black channel is still present in the FST sample, but not in the TERT.

Above my station, but our expert also said that the likelihood of all photon values being /this/ similar is.... "unlikely".

Finally, in this FISH image, you can *clearly see* that the control and treatment are the same cells from the same plate, shot at slightly different exposures.

There's also another suspected image manipulation hiding in there that for boring reasons will remain undisclosed for now.

I expect this level of dodgy science in papers by fringe outfits published in low rent predatory journals. But it is weird seeing it in PNAS with Church's name attached.

It’s possible that there is an innocent explanation for all of this. I emailed Church with my concerns but received no reply. I'd like to think he just signed off and didn't look too closely (apparently neither did PNAS) but Church has presented the paper on at least one occasion

Which all leads back to what's been bothering me: what hold has Parrish/BioViva got over Church, and why is such an esteemed scientist sticking with such dubious company? It's a mystery!

Anyway, if you want to follow me into the rabbit hole, Buying Time is now available on Audible audible.co.uk/buyingtime

a spur, as I forgot to say: I don't believe Church was involved in carrying out this research or preparing the images. It's just weird he's involved at all.