[bonsai_spool]

> very had to predict what binds to what, which of course is a multi-billion-dollar industry.

Do you need to predict when AP-MS is so cheap?

Mapping interaction interfaces is challenging and is where there is attention. I don’t think we’re going to get complexes as a commercial focus outside of receptors with known quaternary structure. The first issue, as you allude to, is the absence of training data, which itself highlights the relative commercial unimportance of such an endeavor.

[chromatin]

> Do you need to predict when AP-MS is so cheap?

Yes because a core reason fro this technology is evaluation of de-novo designs (small peptide binders; scFvs; sdAbs)

[bonsai_spool]

> Yes because a core reason fro this technology is evaluation of de-novo designs (small peptide binders; scFvs; sdAbs)

You are not understanding the distinction. OP was speaking about macromolecular complexes, which are explicitly not what the current round of tools are good at. This is not the same as peptides, small molecules or antibody binding.

[margalabargala]

> Do you need to predict when AP-MS is so cheap?

Yes, because the expensive part is making the thing.

[bonsai_spool]

I responded to a sibling: https://news.ycombinator.com/item?id=48440267

> Yes, because the expensive part is making the thing.

I'm not sure what 'making the thing' refers to here. We are very far from creating designer macromolecular complexes outside of very contrived settings—and there's not a clear reason to focus on getting better than we already are.