> Are they looking for antibodies that attach to specific features unique to SARS-CoV-2?
They're all slightly different, but yes they're looking for antibodies against specific parts of SARS-CoV-2, like the N protein [0][1]. I think the N protein ones are most common. I just did a BLASTp against SARS-CoV's N protein and there's maybe ~90% homology? So I would hope they're using a site that's different between the two. Or, there's an assumption that most people have not been previously exposed to SARS-CoV or others with similar N proteins.
> Also aren't antibodies effectively developed in a sort of random process?
Yeah, but there's only so many prominent features to a virus that you can make antibodies against.
They're all slightly different, but yes they're looking for antibodies against specific parts of SARS-CoV-2, like the N protein [0][1]. I think the N protein ones are most common. I just did a BLASTp against SARS-CoV's N protein and there's maybe ~90% homology? So I would hope they're using a site that's different between the two. Or, there's an assumption that most people have not been previously exposed to SARS-CoV or others with similar N proteins.
> Also aren't antibodies effectively developed in a sort of random process?
Yeah, but there's only so many prominent features to a virus that you can make antibodies against.
[0] https://www.abcam.com/novel-coronavirus-igg-antibody-detecti... [1] https://www.ncbi.nlm.nih.gov/protein/QHW06046.1?report=fasta