News for hackers. Stuff that matters.
Re:Chemical space is really big (2014) by krab on Friday June 25, 2021 @04:50PM (#27636094)
I think that the biggest issue isn't the chemical space but the complexity of biological systems. It's hard to tell what the molecule will do. We just don't have a good enough simulator. AlphaFold is definitely a helpful step but more are needed in the same direction.

(In 2010 - 2012 I worked in a laboratory that did small compounds screening and I was building some tools to explore the chemical space)

Re:Chemical space is really big (2014) by derefr on Friday June 25, 2021 @07:13PM (#27637387)
That’s specifically about chemicals presumed for physiological use/purpose, though, no?

Let’s take biology out of the equation. Can we predict chemical structures will give rise to interesting physical properties when you have a large amount of the compound around?

For some examples, could we ever potentially have a formula or model that would allow us to predict:

• whether any other molecules will possess the “infectious amalgamation” effect that gallium has on other metals it touches?

• what chemicals would create a “non-stick” surface like Teflon does?

• what types of chemicals would be both strongly colored in the visible spectrum and also chemically stable, i.e. what chemicals would make especially good pigments?

• whether something will be a room-temperature superconductor?

• whether something would make a better non-reactive glassware than borosilicate glass?

• what types of chemicals will be especially good at being electrolytes?

And I’d especially like to know, whether there’s some property that unites all “chemicals with interesting physical properties” like this, i.e. whether it’d be possible to have a model that tells you that a chemical is likely to have unusual physical properties in some way, despite not being able to predict the specific effect it’ll have.