My former colleagues at Novartis just signed a $1.7B deal with a seed-stage company.

My former colleagues at Novartis just signed a $1.7B deal with a seed-stage company.

For targets. Not for a drug. Not even for a molecule.

Only a few years ago, target discovery was a cost center, something pharma did internally before the "real" work began. Nobody paid billions for targets, they paid for clinical-stage assets.

That is starting to shift.

Relation has not even yet raised a Series A. Around $80M in seed money, yet they have now locked in deals with GSK and Novartis worth almost $2B in combined milestones.

The logic is simple and brutal: roughly three out of four drugs entering Phase 2 fail.

Relation's answer to that well-known problem is "lab in the loop." Generate proprietary multi-omic data from patient tissue, use AI to identify causal genes, then validate experimentally before you nominate a target.

It is an ambitious stack, a lot of biology has to work for this to pay off.

But the deal structure already tells you something about where pharma sees risk. They are paying earlier in the value chain, betting that better target selection based in experimental biology will reduce that Phase 2 attrition wall.

DCVC and NVIDIA's nVentures backed Relation early. Now big pharma is writing cheques against the same thesis.

Targets are worth real money again, if there is true experimental validation. That alone is worth paying attention to.