From last time ... there are two methods to test for infection with SARS-CoV-2 (the virus which causes covid-19): look for the RNA or look for antibodies.
(1) looking for RNA ... why it's hard, but very reliable.
So you have a swab, and you've processed it to extract a bunch of RNA and you want to see if any of that RNA matches the approximately 29,000 letters that define the SARS-Cov-2 RNA.
Your swab may have millions of bits of RNA from all kinds of viruses. You might, for example, have a cold; which can be caused by multiple viruses. This is because the name 'cold' is just shorthand for a fuzzy bunch of symptoms. And to make matters worse, your millions of bits of RNA will be in little pieces, not a big continuous string. It's like somebody has cut up a bunch of books with scissors and you are looking at the scraps of paper trying to work out if a particular book was in the bunch.
Suppose you found a piece of paper with this on it: "Rami moves by memory through a straight stretch, and calculates the camber of an upcoming turn." If the book you were looking for is Apeirogon, then you'd be pretty safe in saying that it was one of the sliced and diced books ... because these words are the second sentence in Chapter One. The RNA tests for SARS-Cov-2 work the same. They looks for a few short snippets of RNA that are in SARS-Cov-2 and not in any other virus.
So far so good. Picking the snippets isn't straight forward because most virus particles will be ever so slightly different from each other. Imagine if books weren't printed by copied by people; as they used to be back in the times when monks sat in monasteries copying books. You can be certain that most books would have copying mistakes. In the case of the 29,000 or so letters in SARS-Cov-2, changes are quite common. So the trick is to pick snippets that you know are rarely copied wrongly. If that last sentence bothered you, then it should! It hides a wealth of complexity which we'll ignore for now ... just trust that the people choosing the snippets know what they are doing.
This kind of testing is done in laboratories and takes hours but is precise. If you have the virus it will give a positive result and if you haven't got the virus then it will give a negative result. There is one important caveat here. The swab has to be a good one ... meaning from far enough back in the throat or nose to be sure of having virus on it.
In the next part, I'll take about the antibody tests and the absolutely critical issue of false positives and false negatives.