[gojomo]

I'm a believer in the importance of the microbiome, but this study seems fishy:

* the authors have already started a company and pursued a patent on their ayurvedic-derived formulation

* the figure 1(a) in their paper (https://www.nature.com/articles/s41598-018-25382-z#Fig1) doesn't show a big difference from plain probiotics to their formulation, but does show some oddities in the Y-axis where, for the control group, there were clearly exactly just 10 flies' mortality measured (integer steps down), but apparently many, many more for other cases scaled to the 0-10 axis.

* it's unclear if the treatments/evaluations where blinded – the word 'blind' is not found in the Nature article

* the 1st comment on the Nature article wonders: "the control flies didn't live to their normal expected age or even close. What gives?"

(Perhaps the flies that lived longer just had more food in total?)

[ImaCake]

I'm glad someone else shares my skepticism. Looks like they may be guilty of P-hacking in those graphs.

I think this paper has value, in that it basically replicates the SCFA diet results others have seen. But the qPCR graphs are kinda hopeless.

>the authors have already started a company and pursued a patent on their ayurvedic-derived formulation

The fruit fly microbiome is exactly 2 organisms. The human microbiome is somewhat more complex - several hundred genera at least. It's pretty hard to draw a connection between the microbiomes of these two animals.

[chmike]

I would´t trust my nose to detect fishiness. These expermients simply need to be reproduced independently to be verified. Until then, it's anti-scientific to assume anything about fishiness. The reported facts do NOT PROVE that the results were forged.

[krageon]

Nobody is claiming that it does. They are indicating possible economic investment in the results and some questionable statistics. It is perfectly scientific to take these pieces of evidence and adjust the possibility of this study being fishy upwards.

[chmike]

Claiming the presented result as fishy based on the given argument is equivalent to claiming the data has been forged, the scientists are dishonnest and have unethical behavior. Unless the results are proven wrong, such claims of fishyness is defamatory. This is what I call judging by its nose. It's not because such behavior has become common place that it is correct and acceptable.

It would make a huge difference if the OP simply called to be cautious for the given reason. This preserves the status of unknown to the validity of the reported fact without taking position one way or the other.

It is very disapointing that people can't make the difference and I maintain that such behavior is anti-scientific. Claiming the opposit doesn't make your point true.

People who make claims like this fishyness are people who have a very high opinion on their ability to distinguish true from false facts just by guessing. That's monkey science.

[gojomo]

'Fishy' doesn't mean 'forged', just 'suspicious'.

The factors I mention – fast patenting/profit-seeking, odd axes/combinations to get a result, unclear blinding, inconsistency with other expected fly lifetimes – are all the kinds of things correlated with flimsy results. It's usually wishful thinking, not conscious forgery, that leads such authors to overlook the weaknesses in their setup when they get a publishable/profitable result.

I do have a high opinion of my ability to detect flimsy results from the details (or missing details!) of a scientific paper, from decades of reading and watching which results hold up, and which don't.

You're practicing scientism – sacralizing certain procedures, titles, or outlets – rather than science here. Science requires a high standard of proof, and recognizing patterns of misreporting. There's even a strong case to be made that "most published research findings are false":

http://journals.plos.org/plosmedicine/article?id=10.1371/jou...