Luke O'Neill

And it carried a poison with it.

They're called antibody drug conjugates.

So you attach a nasty chemical to an antibody that targets this thing, B7H4, and the chemical kills the cell.

There's been a few ideas around this already in the past.

This is probably the first example of it really working.

I mean, you see, that kind of response would predict a huge effect down the line, then, is the idea.

And what was interesting about it, they got it from China.

Increasingly, the Chinese are discovering new medicines.

This wasn't the case in the past.

And GSK saw this Chinese company, looked at the drug, looked at their data, bought the drug in from China, and now we're seeing this big effect on ovarian.

And GSK themselves are saying this will be a blockbuster.

Because that ovarian is very common.

And again, not much is about to treat it.

If you can imagine a drug that gets through, they're predicting billions in sales.

How much of this is to do with AI being employed, or is this just old-fashioned, good, solid research?

I think these are good-fashioned, old-fashioned stuff at the moment, in a way.

They will use AI path on the trials, remember, to assess the patients and see who's responding.

And I suspect with GSK, they found that 62% response rate through AI probably is part of it.

Okay, so it seems like everything's gathering apace.

I mean, I see stuff online every day about development and Parkinson's disease and all of that.