Comment by _Microft | Hacker News Reader

Comment by _Microft | Hacker News Reader

Parent Full thread

Here is the paper [0] and an attempt at a summary:

A characteristic of Parkinson's disease ("PD") [1] seems to be that there are lumps [2] of a protein (alpha-synuclein, "alpha-syn" [2a]) in the neurons of patients.

What they seemed to have done is: they isolated "Desulfovibrio" bacteria strains from the feces of both Parkinson's disease patients and their healthy partners. Then they fed these bacteria to worms (nematodes [3], in particular [3a]) that were genetically modified to produce more alpha-synuclein which itself was modified to connect to green fluorescent protein (GFP [4]). They imaged the heads of the worms to detect clusters of green fluorescence.

Worms that received bacteria from PD patients showed more of these clusters than the control groups of which one was fed bacteria from healthy subjects and the other just with E.coli bacteria.

I hope I got that right.

[0] https://www.frontiersin.org/articles/10.3389/fcimb.2023.1181...

[1] https://en.wikipedia.org/wiki/Parkinson%27s_disease

[2] https://en.wikipedia.org/wiki/Lewy_body

[2a] https://en.wikipedia.org/wiki/Alpha-synuclein

[3] https://en.wikipedia.org/wiki/Nematode

[3a] https://en.wikipedia.org/wiki/Caenorhabditis_elegans

[4] https://en.wikipedia.org/wiki/Green_fluorescent_protein

It's a breakthrough but this research doesn't establish the claim in the title of the article.

The next step is to establish causality. They will have to deliberately remove (and maintain removal of) this bacteria from certain people and measure the progress of parkinsons over many years. There will also be a control group given placebos. The difference between these two groups will be the factor that crystallizes the claim made here.

It's even worse. This study was done in worms. A species of worms widely known for having literally one of the most simple neural structures of any animal on the planet (they have like 300 Neurons in total). Those worms can't even develop Parkinson's. So what they observed is that literally feeding the bacterium in large quantities to those worms coincides with buildup of a certain protein that regulates synaptic activity and whose build up is observed in a certain share of humans with Parkinson's. But the worms were also genetically modified to enhance their susceptibility towards that protein. So this result doesn't just have to jump the causality hoop, it first has to make the jump to independent confirmation. Because there is a lot of potential for unintended bias. Even if it manages that, it also has to make the jump to mammals and after that it could still fail in the clinical stage. It is at best news in a field deprived of breakthroughs for a long time, but it is far from a real breakthrough.

I only read a summary, but that is not how I interpreted the use of the worms. My understanding was that they were using the worms as a sort of biological measuring device to compare levels of that particular chemical. Is that not the case?

> It's a breakthrough but this research doesn't establish the claim in the title of the article.

What claim would that be? The only claims made in the article was that a) it's a breakthrough, and b) "Researchers say certain strains of gut bacteria are the likely cause of Parkinson's disease."

The article also refers to the bacterias as "probable causes".

Everything you pointed out supports these observations.

“Associated with” would have been a stronger claim than “probable cause of”.

> “Associated with” would have been a stronger claim

in English, "stronger claim" has a meaning ambiguous between "the claim is strongly defendable, i.e. more likely to be true", and "what is being claimed is of more powerful effect". I think here the claim they are trying to point out is the powerful effect, to say "this should be pursued, its importance is elevated by its potential strong causal/explanatory effect"

"probable cause" means the same thing as "associated with", but establishes a specific hypothesis and stresses the fact that it is yet to be verified.

No. Causation is different from correlation. "Associated with" means correlation, "probable cause" means causation. Causation is the stronger statement here.

While technically association is a prerequisite for causation (correlation doesn't imply causation but causation implies correlation) probable "cause" is too strong of a term here because the experiment didn't actually do a causative test. "Possible cause" might be better here because the title heavily implies a causative study was done.

Do note that in science nothing can be proven so in actuality probable cause is really the highest form of verification that can be made. That is the claim in the title.

Reply on news.ycombinator.com