So if we think now about the digital factor here, what role can digital twins play here, Irina? So

digital twins is a concept that has been used in other industries for a very long time, which means it's a computational simulation of a physical system and you can run multiple simulations and multiple scenarios before you realise it in the physical world.

And engineering and aerospace engineering in particular is where it's been used for a very long time.

In medicine, we have actually been adopting it for a while with processes and with different other systems, but it hadn't quite penetrated biology because inherently we don't understand biology.

We don't really know how cancer works.

All the data about cancer is highly fragmented and it's all over the place in different formats.

So.

I mean, with this or these regulatory issues remaining, what additional hurdles alongside that consideration of those frameworks remain, Irina?

Why is prediction so hard?

Well, it is because, um, of the very thing you said.

We generate a lot of the data.

We generate different types of data, and each one of these data sets kind of carries, uh, a little bit of information about biological state.

But integrating all of that knowledge together is a challenge because these are really disparate data sets.

They are often generated in very different ways.

So to actually be able to integrate and share learnings from all these data sets together in order to get a more complete picture of cancer biology has been challenging with, uh, conventional bioinformatics approaches and conventional data, um, analytics until recently, of course, where we can do a lot of da-data processing really quickly.

Okay