Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

Most of the comment is correct, but

> Or maybe we could test how the proteins in the sample chemically react.

> ... but that doesn't help, because the atoms are connected the same way.

When the shape of a protein changes, the chemical reactions may change too. For example some proteins usualy bind to a specific molecule, but if you twist it incorrectly the "hole" where the molecule binds is broken. So it's theoreticaly possible to check this, but I don't know if it's easy in a lab.

> The only difference is in how the proteins are twisted...

> ... and we can't even figure out how proteins are supposed to be twisted in the first place,

In many important proteins we know.

> let alone how they're twisted in a specific sample.

From a sample there are sote techniques like x-ray difraction, but you need very pure samples, and be lucky that the protein forma a nice crystal, and perhaps a lot of other problems. We have done these in the past to determine the proteins shape before computers, but I think it's just too difficult with a small sample of human tissue.



Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: