4 ms·

The thought of an as-yet-unknown CJD-like prion disease that's slowly developing somewhere is absolutely terrifying. This made me wonder about how tests are run

by bink 5y ago

The thought of an as-yet-unknown CJD-like prion disease that's slowly developing somewhere is absolutely terrifying. This made me wonder about how tests are run for prion diseases, though. Is it just a matter of placing the spinal fluid sample under a microscope or is it more complicated than that?

supernova87a 5y ago
From my 2nd-hand info (my comment below about friend-of-family stricken with the disease), I understand that the local hospital was unequipped to diagnose it fully. They had to send a sample to CDC (?) or maybe Mayo as part of autopsy. If I remember what I read years ago (see "Deadly Feasts" book by Richard Rhodes, I recall it being a good read, and part of what made me swear off beef) they actually look at the brain matter to see if the amyloid plaques are present -- the degenerated holes in the brain tissue. I didn't think there was some molecular-kind of test for the proteins themselves. But my info is from years ago on this.
- bink 5y ago
  The article says they took spinal taps from the patients and found no evidence of CJD. Presumably that means it can be detected that way, but how? And is it close to 100% reliable?
  - abeyer 5y ago
    https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5266667/ https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5266667/
ampdepolymerase 5y ago
The most effective thing you can do as a software engineer is to contribute to applied machine learning research and implementation in bio/chemoinfomatics. Solve protein folding, and entire classes of such diseases would be curable. DeepMind has given computational biology its ImageNet moment, but there is still a large amount of room to improve for production level use (beyond merely as a research tool). Truly talented ML engineers willing to work on biology are very rare.
- carlmr 5y ago
  >Truly talented ML engineers willing to work on biology are very rare. I think it's mostly that in the biology space, at least in my country, they will be paid less than half of a software engineer's salary.
  - elric 5y ago
    And they're often paid based on education level, which is pretty silly at the best of times. Many software engineers have no degree, if they want to work in academia/research labs, they're often unhirable.
jamespwilliams 5y ago
CJD is almost always a clinical diagnosis. MRIs and lumbar punctures give hints, but aren’t conclusive. The only way to diagnose conclusively is via a brain biopsy, which apparently isn’t worth the risk in most cases. Conclusive diagnosis is often done post-mortem. (my girlfriend works in this area)
throwawayboise 5y ago
Why is it absolutely terrifying in comparison to the many other more commonplace diseases that can kill you at any time?
- supernova87a 5y ago
  Agree with other comment. It's scary because it has no way to be destroyed, aside from incineration of items you know are contaminated (which often people don't know are contaiminated), or possibly dilution in the environment to undetectable levels. And aside from that, the progression of the disease and the neurological symptoms/effects on a person are terrible.
- Auracle 5y ago
  We have absolutely no way to treat prion diseases. It’s a death sentence - plus anything neurodegenerative is scary.
- lovemenot 5y ago
  Such diseases are different. Prions are not even at virus-level of livingness. Like a propogating hardware bug. Or like Vonnegut's ice-nine. I find it interesting that at least the three I am aware of: scrapie, BSE and kirii - maybe others too - are all associated with cannibalism. In sheep, cows and humans respectively.
  - fourtrees 5y ago
    Kirii? I don't mean to be pedantic, but do you mean kuru-kuru? Or is there another prion disease out there I'll be having nightmares about?
    - lovemenot 5y ago
      Thanks. You are right. I misremembered the name. Shame on me since I actually once met an old woman in Papua New Guinea who had participated in funerary cannibalism as a young girl. She had had no ill-effects though.
- 5y ago

by bink 5y ago

supernova87a 5y ago

From my 2nd-hand info (my comment below about friend-of-family stricken with the disease), I understand that the local hospital was unequipped to diagnose it fully. They had to send a sample to CDC (?) or maybe Mayo as part of autopsy. If I remember what I read years ago (see "Deadly Feasts" book by Richard Rhodes, I recall it being a good read, and part of what made me swear off beef) they actually look at the brain matter to see if the amyloid plaques are present -- the degenerated holes in the brain tissue. I didn't think there was some molecular-kind of test for the proteins themselves. But my info is from years ago on this.

bink 5y ago
The article says they took spinal taps from the patients and found no evidence of CJD. Presumably that means it can be detected that way, but how? And is it close to 100% reliable?
- abeyer 5y ago
  https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5266667/ https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5266667/

ampdepolymerase 5y ago

The most effective thing you can do as a software engineer is to contribute to applied machine learning research and implementation in bio/chemoinfomatics. Solve protein folding, and entire classes of such diseases would be curable. DeepMind has given computational biology its ImageNet moment, but there is still a large amount of room to improve for production level use (beyond merely as a research tool). Truly talented ML engineers willing to work on biology are very rare.

carlmr 5y ago
>Truly talented ML engineers willing to work on biology are very rare. I think it's mostly that in the biology space, at least in my country, they will be paid less than half of a software engineer's salary.
- elric 5y ago
  And they're often paid based on education level, which is pretty silly at the best of times. Many software engineers have no degree, if they want to work in academia/research labs, they're often unhirable.

jamespwilliams 5y ago

CJD is almost always a clinical diagnosis. MRIs and lumbar punctures give hints, but aren’t conclusive. The only way to diagnose conclusively is via a brain biopsy, which apparently isn’t worth the risk in most cases. Conclusive diagnosis is often done post-mortem. (my girlfriend works in this area)

throwawayboise 5y ago

Why is it absolutely terrifying in comparison to the many other more commonplace diseases that can kill you at any time?

supernova87a 5y ago
Agree with other comment. It's scary because it has no way to be destroyed, aside from incineration of items you know are contaminated (which often people don't know are contaiminated), or possibly dilution in the environment to undetectable levels. And aside from that, the progression of the disease and the neurological symptoms/effects on a person are terrible.
Auracle 5y ago
We have absolutely no way to treat prion diseases. It’s a death sentence - plus anything neurodegenerative is scary.
lovemenot 5y ago
Such diseases are different. Prions are not even at virus-level of livingness. Like a propogating hardware bug. Or like Vonnegut's ice-nine. I find it interesting that at least the three I am aware of: scrapie, BSE and kirii - maybe others too - are all associated with cannibalism. In sheep, cows and humans respectively.
- fourtrees 5y ago
  Kirii? I don't mean to be pedantic, but do you mean kuru-kuru? Or is there another prion disease out there I'll be having nightmares about?
  - lovemenot 5y ago
    Thanks. You are right. I misremembered the name. Shame on me since I actually once met an old woman in Papua New Guinea who had participated in funerary cannibalism as a young girl. She had had no ill-effects though.
5y ago