They did correct for multiple testing - "P-values were multiple-test corrected using the Benjamini-Hochberg False Discovery Rate (FDR) procedure and a cutoff of 0.1 was applied to select the differentially abundant genera". However, they used Wilcoxon test which has unacceptably high FDR [1].

[1] https://www.nature.com/articles/s41467-022-28034-z

Thanks for that great write-up. I wonder if there are microbiome studies with actually large effect sizes in diseases which we do not typically associate with infectious factors (e.g. diabetes).

That doesn't solve the problem of funneling public money to elsevier/springer nature. The model in which volunteers perform 95% of editorial work yet private publishing companies charge abhorrent prices for OA is absolutely disgusting.

Well, the reason for the hate could be that OR things like oil companies sponsoring climate change deniers and exxon hiding results of their own study which shown that continued burning of fossil fuels will have catastrophic consequences.

What? Why?

The flagged post didn't put it eloquently but the sentiment is right, and complaining about naming violates HN rules about not complaining about tangential issues. As it is they've at least temporarily ruined the discussion by having the top post be feigned concern about something that's not at all material to the content.

>Data scientist here who spent a couple of years working with Unreal (to produce high end data visualizations) That sounds amazing, could you share some of your visualizations?

How to learn it?

> Cryptococcal meningitis is the most common cause of central nervous system infection in people living with HIV worldwide. Isn't HIV encephalitis more common?

Encephalitis is inflammation of the brain. It can be caused by infection but the word does not actually describe infection per se.

I think the commenter is referring to acquired HIV encephalitis caused by direct HIV infection which I was also under the impression was the #1 CNS infection in HIV patients.

My teaching (in radiology) was HIV > toxoplasmosis > cryptococcosis for CNS infections in HIV+ but maybe we're out of date or this order is sepcific to the US/Canadian population.

Anecdotally I've definitely seen more toxo than crypto. I've also seen more white matter disease in HIV patients than either but the MRI findings aren't specific so I don't know what the final path was on those cases.

In our HIV patients we see more more cryptococcal meningitis than CNS toxoplasmosis. Crypto is typically not going to have any significant radiologic abnormality unlike toxo in which imaging plays a large role in diagnosis. So I'm guessing, being a radiologist, you've got a sampling bias that favors toxo.

PO amphotericin B would be a huge boon in treating these patients and shortening hospital stays. Outpatient Ampho B is not a good option in most cases.

Just in case it's not clear I am by no means claiming domain expertise, merely stating that what I was taught and my understanding was similar to the initial comment I replied to hence the caveats and soft language. My statement should not be read as contradicting an ID expert or claiming that the author of the article is incorrect.

> In our HIV patients we see more more cryptococcal meningitis than CNS toxoplasmosis. Crypto is typically not going to have any significant radiologic abnormality unlike toxo in which imaging plays a large role in diagnosis. So I'm guessing, being a radiologist, you've got a sampling bias that favors toxo.

Agree crypto is much more subtle on imaging than either HIV encephalitis or toxo, the most common finding we see is dilated PVS which is nonspecific (particularly without priors). I only mentioned my anecdotal experience as it corresponds with what's taught to us but I agree it's highly susceptible to bias and I don't consider it evidence.

For example on StatDx (UpToDate for radiologists):

>[Cryptococcus is the] most common fungal infection in AIDS patients

>3rd most common [CNS] infection seen in AIDS patients (HIV > toxoplasmosis > Cryptococcus)

This could very well be out of date/incorrect, they don't give in-text citations like UpToDate so I'm not sure where these specific statements are coming from.

Do you have a reference handy? If so I can submit it as feedback on the article to get it updated/reviewed.

Honestly there is conflicting information about which is more prevalent (toxo or crypto). From what I've found the sources that site toxoplasmosis as most common are older, and the ones reporting cryptococcal meningitis as more common are more recent. I suppose the incidence may have shifted since the 90s. I don't really know. Anecdotally I see more crypto (private practice ID in southeast US).

Interesting. Probably did shift then, it would fit the pattern of epidemiological changes taking a while to percolate to radiology and as it's far more likely we miss crypto on MRI than toxo we probably wouldn't notice a change in our reporting incidence to make a radiologist question that ranking.

Thanks for taking the time to search and comment. Always appreciate learning from my clinical colleagues + now I can flex a new obscure fact to radiology trainees like a proper academic physician.

So they finally jazzed it up?

What if it REALLY is too expensive? You do realize that there are studies which literally cost millions of dollars? Getting funding for original studies is hard enough, good luck securing additional funds for replication.