3 ms·

Perhaps a basic question, but what is actually required to test for COVID-19? I've seen mentions of "testing kits", but what's actually in them? For comparison,

Perhaps a basic question, but what is actually required to test for COVID-19? I've seen mentions of "testing kits", but what's actually in them? For comparison, US medical facilities regularly test for any number of viruses and can essentially turn them around indefinitely. Is this because they know they'll continue to need to test for them, so they are well-stocked in the test kits?

singularity2001 7y ago
The most sensitive test is polymerase chain reaction PCR, which multiplies tiny remnants of DNA of the virus It can take up to a day to get results, a team of four can only do roughly about 100 tests per day Then you have serology test, which checks for antibodies against the virus, The signal is strongest after the peak of the disease The Koreans are working on tests which can be applied at home but I don’t know details
- thedance 7y ago
  In 1990's I worked on the software for an automated NASBA amp. These were mostly sold into the market for HIV testing. They are a lot more productive than 25/day/tech. Is PCR the only way to test for this virus? Are patents standing between us and alternatives?
- usaar333 7y ago
  nit, but it's an actually amplifying the RNA (it's an RNA, not DNA based virus), using RT-PCR (which basically converts the RNA into complementary DNA -- and then amplifying DNA) I'm not sure per se what the CDC does, but here's a good overview of a testing method: https://www.broadinstitute.org/files/publications/special/COVID-19%20detection%20(updated).pdf https://www.broadinstitute.org/files/publications/special/CO...
bionhoward 7y ago
The machines and reagents are super common. The critical missing element is the primer, a short sequence of nucleotides telling the PCR (gene copier) what sequence to copy Gene synthesis companies are probably ramping production of these primers now. It’s a logistics issue to send them to the labs, but tiny quantities are needed so the packages are gonna be small, with cold chain shipping I think. We (molecular biologists) have a dearth of tools for affordable label-free testing, so you basically have to know what you’re looking for before you can look. Bad problem to have. this could be improved with sequencers from Illumina / Roche / Oxford Nanopore but need $$ for that

singularity2001 7y ago

The most sensitive test is polymerase chain reaction PCR, which multiplies tiny remnants of DNA of the virus It can take up to a day to get results, a team of four can only do roughly about 100 tests per day Then you have serology test, which checks for antibodies against the virus, The signal is strongest after the peak of the disease The Koreans are working on tests which can be applied at home but I don’t know details

thedance 7y ago
In 1990's I worked on the software for an automated NASBA amp. These were mostly sold into the market for HIV testing. They are a lot more productive than 25/day/tech. Is PCR the only way to test for this virus? Are patents standing between us and alternatives?
usaar333 7y ago
nit, but it's an actually amplifying the RNA (it's an RNA, not DNA based virus), using RT-PCR (which basically converts the RNA into complementary DNA -- and then amplifying DNA) I'm not sure per se what the CDC does, but here's a good overview of a testing method: https://www.broadinstitute.org/files/publications/special/COVID-19%20detection%20(updated).pdf https://www.broadinstitute.org/files/publications/special/CO...

bionhoward 7y ago

The machines and reagents are super common. The critical missing element is the primer, a short sequence of nucleotides telling the PCR (gene copier) what sequence to copy Gene synthesis companies are probably ramping production of these primers now. It’s a logistics issue to send them to the labs, but tiny quantities are needed so the packages are gonna be small, with cold chain shipping I think. We (molecular biologists) have a dearth of tools for affordable label-free testing, so you basically have to know what you’re looking for before you can look. Bad problem to have. this could be improved with sequencers from Illumina / Roche / Oxford Nanopore but need $$ for that