Thread by @davidasinclair

Here’s how PCR tests work, why they are flawed, and why people are retesting positive...

https://twitter.com/davidasinclair/status/1250986393221947393

First: how #Covid_19 PCR tests work. It’s 1980s tech, a kid could run one, but if you don’t care, skip down to the flaws...

1. Get a nose or lung mucus/cell sample, deeper the better.

2. Purify the viral RNA and turn it to DNA using reverse transcriptase (RT), hence it’s name, the “RT-PCR” test.

3. Add in two types of short pieces of DNA called “primers” that stick to precise places on the viral genome. Because DNA is a 2 stranded zipper (with strands running in reverse directions), the primers are made to bind to opposite strands, and point towards each other.

Tidbit 1: the CDC screwed up the design of the DNA primers. Bad design leads to them sticking to other places in the human genome or failing to bind efficiently to the virus.

Tidbit 2: Primers are cheap these days. You can buy a quintillion of them for $1. That’s the number of different ways a Rubik’s cube can be arranged.

4. The tiny reaction also has a heat-stable DNA polymerase, which acts like that little handle on a zipper, but it’s special. It grabs onto the primer and, while it’s closing it, makes a whole new side of the zipper.

5. After a few minutes, the DNA zippers are heated to 203F to separate the DNA zipper stands. 🧬 When the reaction cools, the process is repeated, amplifying the DNA exponentially. After 30 cycles you have about a billion copies of each viral genome.

6. #COVID tests use real-time quantitative PCR (qPCR), which have fluorescent dyes in the mix to tell the bench-top machine how much DNA is generated.

7. The Ct value is the PCR cycle number at which the sample’s fluorescence passes a predetermined threshold line. Now you can read those graphs.

8. PCR tests are so sensitive they can theoretically pick up a single #coronavirus particle. As we will see, this is good news and bad...

Now that we know how PCR works, let’s consider the headlines trying to make us worry about #corona reinfection... tinyurl.com/y8hfp5cb

1. When Korean scientists tried to culture #coronavirus from people retesting positive, they couldn’t. The PCR tests are likely detecting non-transmissible, residual viral RNA.

2. A paper out of China by Zhuang et al. found “in the close contacts of #Covid_19 patients, nearly half or even more of the ‘asymptomatic infected individuals’...might be false positives.”

3. This would explain how numerous members of a Korean family retested positive: they were probably contaminated with RNA from a family member (or their home is covered in it). It’s also possible their earlier tests failed.

The PCR tests for #coronavirus have major flaws:

1. They are exquisitely sensitive and can pick up environmental contaminants. Labs doing the #Covid_19 tests may be covered in #coronavirus.

2. PCR tests are not binary (negative/positive). They have an arbitrary threshold, the Ct.

3. Results outside the threshold (e.g. above 38 cycles of the PCR) are treated as negative, but if PCR continues past the cutoff it may be positive.

4. Negative doesn’t always mean uninfected and positive doesn’t always mean infected.

5. Sampling is never flawless either. Swabbing for 10 seconds in the far recesses of the nose where the virus likes to incubate is painful and difficult. It’s easy to give up. #flawedtests

Someone whose tests failed could later come back positive if the sampler knows how to get a good sample.

We always develop immunity to seasonal viruses like flu & colds by making IgG antibodies that seek and neutralize viruses. If we didn’t, we’d be extinct. My bet is the tests are flawed, not the human immune system.

Conclusion: it is extremely unlikely people are going from infected to uninfected and back again, sometimes multiple times.

The antibody tests are different. They are designed to tell if you’ve got #Covid_19 antibodies or immunity. They are also flawed. Like this thread a lot and I may write about those too ✌️