I'm curious to what extent this sort of data depends on positive tests. It seems there's kind of a contradiction here: we can detect huge numbers of mutations,

I'm curious to what extent this sort of data depends on positive tests. It seems there's kind of a contradiction here: we can detect huge numbers of mutations, but only if someone tests positive as otherwise a sample will be ignored by the labs (presumably?). Yet the tests are supposed to be extremely precise. Does anyone know how the need to detect mutations is balanced against the need for the tests to be precise to a specific type of virus?