[charli (OP)]

Pete has been wondering:

"How are new viruses discovered when you don't know what you are looking for?

Can the system we have for discovering new viruses be automated? Could blood samples from sick patients around the world be randomly selected for further analysis and any new viruses uploaded automatically to a repository for comparison?"

Can you help him out?

[evan_au]

You can look under an optical or electron microscope, to see if you can see anything that looks out of place.
- Then compare it to the shape of known pathogens

There is a technique where you just use modern high-speed DNA sequencing to read every piece of DNA and RNA in a sample. Then you compare the genetic sequences to a library of known DNA sequences.
- Many sequences will come from the human or animal host, so you can discard those
- Many sequences will come from known, benign bacteria, so you can discard those
- You can compare the "unknown" sequences from different patients who seem to have the same symptoms, and see if the unknown DNA sequences match up. A match indicates that you have discovered a potential pathogen
- Often, these sequences will have general similarities to other pathogens already on file, so you can tell whether it is a virus or a bacteria, and which general family they belong to.

There was a project to just read the genetic sequence of sea-water. This revealed many unknown varieties of microbes, as well as taking a census of the expected fish...
See: https://www.illumina.com/company/news-center/feature-articles/sequencing-the-seas.html

[CliffordK]

For diseases, one can look for a novel presentation of illnesses, or a particularly severe outbreak.  One problem is people who either don't see a physician, or each person goes to a different physician, so good community communication is important.