Ta. It‘s a very difficult area to read critically as a non-expert. I did like the use of FISH/IHC though, even if we don’t know what bacterial presence actually means.

I’m not a metagenomics person, but I’m thinking: methodological issues are the lack of a kit control, and lots of samples with no signal; and then interpretation of clinical relevance / causal relationship is really hard anyway. Am I going along the right lines?

The aim is twofold: dogs could be used directly to triage people with suspected infection, but they also generate high-quality data sets to train the AI that powers electronic noses.

How about this one? futurism.com/artificial-i...

Lovely work by @c-cornbill.bsky.social @kadeayomikun.bsky.social @nreddy.bsky.social @microbescience.bsky.social @tosinorababa.bsky.social, pooling their skills with lung, wound and endotracheal tube models to explore the potential of tetrasodium EDTA in different infection contexts #MicroSky

Thank you for getting in touch, I’ll email tomorrow 🙂