4 ms·

Kary Mullis also doubts that HIV causes AIDS, believes in astrology, and once saw a fluorescent talking raccoon. Clearly a brilliant guy, but not someone whose

Kary Mullis also doubts that HIV causes AIDS, believes in astrology, and once saw a fluorescent talking raccoon. Clearly a brilliant guy, but not someone whose advice should directly determine clinical practice. https://www.mcgill.ca/oss/article/technology-history/man-who-photocopied-dna-and-also-saw-talking-fluorescent-raccoon https://www.mcgill.ca/oss/article/technology-history/man-who... See my other comment here for the reasons why I believe almost all SARS-CoV-2 PCR positives are true positives (as to whether the patient is or was very recently infected; the kernel of truth is that they don't necessarily indicate that the patient could infect others right now).

silicon2401 5y ago
That's an ad-hominem argument. You might as well say Isaac Newton also believed in astrology, so we should throw out all his work on math and physics. As a non-religious person, astrology and religion are no different to me, so then you might as well throw out any knowledge innovated by like 99% of people who have ever lived. I have no stake or attachment to Kary Mullis, whether he's wrong or right is irrelevant to me. I just find it interesting that the creator of PCR, which has a lot of attention these days, had opinions that don't get a lot of attention these days. I'll give your comment a look, but am also curious if anybody else has a counter to Mullis' claims.
- tripletao 5y ago
  > That's an ad-hominem argument. And yours was an argument from authority. You gave no explanation of or support for Mullis's arguments, and you gave no reason to believe his conclusion except a positive aspect of his reputation (i.e., his discovery of PCR in the first place). In that case, negative aspects of his reputation (i.e., the fluorescent raccoon; or, closer to the domain in question, his belief on HIV and AIDS) are surely also relevant. They don't mean what he's saying is false, but they'd cause a reasonable person to approach his unsupported statements with more care. For convenience, I'm linking my other comment below. This is a detailed explanation of why I believe almost all patients who test positive by PCR are (or were very recently) infected with SARS-CoV-2, fully referenced. It has so far received zero interaction. https://news.ycombinator.com/item?id=26984429 https://news.ycombinator.com/item?id=26984429 If you think I'm wrong, please explain why, and I'll reply. If you think I'm right, then I'd also appreciate a reply. The belief that coronavirus cases are grossly over-diagnosed is a common element of claims that no precautions against it are necessary, which cause real harm; so I hope you don't want to falsely spread that belief. ETA: And for completeness, here's a paper discussing the ability to culture virus from PCR-positive samples. I believe that's the real point Mullis was alluding to in his comments, that people who test PCR positive aren't necessarily shedding replication-competent virus; but that's a separate question from whether they are (or very recently were) infected. https://academic.oup.com/cid/advance-article/doi/10.1093/cid/ciaa1491/5912603 https://academic.oup.com/cid/advance-article/doi/10.1093/cid...
  - theclap 5y ago
    You might find this interesting, an article from 2007 about PCR tests being used to confirm a whooping cough epidemic at a hospital: https://archive.ph/ugiWQ https://archive.ph/ugiWQ I wonder if the same mistakes are being repeated.
    - tripletao 5y ago
      From the article: > These tests, called “home brews,” are not commercially available, and there are no good estimates of their error rates. So this is about clinical use of an uncharacterized test. Why is that relevant to the test for SARS-CoV-2, which has probably now been run more times than any other PCR test in history, including on large populations with extremely low test positivity (e.g., most of Australia, <0.1%), which bounds the specificity (e.g., to >99.9%)? Seriously, engage with the evidence. There are still zero replies to my detailed explanation linked above. Did you read it? What did you think?
      - theclap 5y ago
        You didn't read the article. You stopped reading right at that sentence didn't you?
        tripletao 5y ago
        I read to the end of the article. They even developed a better (though still not perfect) PCR test that excluded most of the initial false positives! > Its scientists also did additional P.C.R. tests on samples from 116 of the 134 people who were thought to have whooping cough. Only one P.C.R. was positive, but other tests did not show that that person was infected with pertussis bacteria. It's absolutely possible to develop bad PCR tests, and that's why it's necessary to characterize them on large populations to determine their specificity and sensitivity. It seems the Dartmouth team didn't do that before using the test clinically, and that was their problem; but the SARS-CoV-2 tests are the most intensively-studied PCR tests in human history, and the evidence I've seen looks quite good. So again, did you read my detailed explanation? Do you understand it? If you're just looking to knowingly spread baseless doubt, there's far more effective places to do it. So why not engage with the actual science?

silicon2401 5y ago

That's an ad-hominem argument. You might as well say Isaac Newton also believed in astrology, so we should throw out all his work on math and physics. As a non-religious person, astrology and religion are no different to me, so then you might as well throw out any knowledge innovated by like 99% of people who have ever lived. I have no stake or attachment to Kary Mullis, whether he's wrong or right is irrelevant to me. I just find it interesting that the creator of PCR, which has a lot of attention these days, had opinions that don't get a lot of attention these days. I'll give your comment a look, but am also curious if anybody else has a counter to Mullis' claims.

tripletao 5y ago
> That's an ad-hominem argument. And yours was an argument from authority. You gave no explanation of or support for Mullis's arguments, and you gave no reason to believe his conclusion except a positive aspect of his reputation (i.e., his discovery of PCR in the first place). In that case, negative aspects of his reputation (i.e., the fluorescent raccoon; or, closer to the domain in question, his belief on HIV and AIDS) are surely also relevant. They don't mean what he's saying is false, but they'd cause a reasonable person to approach his unsupported statements with more care. For convenience, I'm linking my other comment below. This is a detailed explanation of why I believe almost all patients who test positive by PCR are (or were very recently) infected with SARS-CoV-2, fully referenced. It has so far received zero interaction. https://news.ycombinator.com/item?id=26984429 https://news.ycombinator.com/item?id=26984429 If you think I'm wrong, please explain why, and I'll reply. If you think I'm right, then I'd also appreciate a reply. The belief that coronavirus cases are grossly over-diagnosed is a common element of claims that no precautions against it are necessary, which cause real harm; so I hope you don't want to falsely spread that belief. ETA: And for completeness, here's a paper discussing the ability to culture virus from PCR-positive samples. I believe that's the real point Mullis was alluding to in his comments, that people who test PCR positive aren't necessarily shedding replication-competent virus; but that's a separate question from whether they are (or very recently were) infected. https://academic.oup.com/cid/advance-article/doi/10.1093/cid/ciaa1491/5912603 https://academic.oup.com/cid/advance-article/doi/10.1093/cid...
- theclap 5y ago
  You might find this interesting, an article from 2007 about PCR tests being used to confirm a whooping cough epidemic at a hospital: https://archive.ph/ugiWQ https://archive.ph/ugiWQ I wonder if the same mistakes are being repeated.
  - tripletao 5y ago
    From the article: > These tests, called “home brews,” are not commercially available, and there are no good estimates of their error rates. So this is about clinical use of an uncharacterized test. Why is that relevant to the test for SARS-CoV-2, which has probably now been run more times than any other PCR test in history, including on large populations with extremely low test positivity (e.g., most of Australia, <0.1%), which bounds the specificity (e.g., to >99.9%)? Seriously, engage with the evidence. There are still zero replies to my detailed explanation linked above. Did you read it? What did you think?
    - theclap 5y ago
      You didn't read the article. You stopped reading right at that sentence didn't you?
      - tripletao 5y ago
        I read to the end of the article. They even developed a better (though still not perfect) PCR test that excluded most of the initial false positives! > Its scientists also did additional P.C.R. tests on samples from 116 of the 134 people who were thought to have whooping cough. Only one P.C.R. was positive, but other tests did not show that that person was infected with pertussis bacteria. It's absolutely possible to develop bad PCR tests, and that's why it's necessary to characterize them on large populations to determine their specificity and sensitivity. It seems the Dartmouth team didn't do that before using the test clinically, and that was their problem; but the SARS-CoV-2 tests are the most intensively-studied PCR tests in human history, and the evidence I've seen looks quite good. So again, did you read my detailed explanation? Do you understand it? If you're just looking to knowingly spread baseless doubt, there's far more effective places to do it. So why not engage with the actual science?