Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Chemistry Nobel: Computational protein design and protein structure prediction

nobelprize.org

Post reply on HN
Print view View original

281–290 of 343 posts

Re: Chemistry Nobel: Computational protein design and protein structure prediction

#281

Post by BeetleB »

huijzer wrote: post #225
BeetleB wrote: post #176

Earlier quoted context omitted.

The patent clerk guy was almost done with his PhD when he became a patent clerk. Not quite comparable.

Are we talking about Einstein? If I remember correctly, according to Walter Isaacson, Einstein managed to get so many good papers out not despite, but because he was not working for an university. It gave him more freedom to reject existing ideas. Also the years I can find on Wikipedia do not seem to support your claim. He started as a clerk in 1903, and had his miracle year and submitted his PhD dissertation in 1905…

From what I've read, he explicitly sought a position that would give him time to work on his physics ideas. Whether he would or not would have achieved the same working for a university is merely his opinion. In particular, it was not the case that he was working for one, and found it to be incompatible with his research vision, and left academia to become a patent clerk.

He began his work on the PhD prior to 1903.

Re: Chemistry Nobel: Computational protein design and protein structure prediction

#282

Post by flobosg »

roughly wrote: post #264
flobosg wrote: post #242

Earlier quoted context omitted.

AlphaFold and Folding@home attempt to solve related, but essentially different, problems. As I already mentioned here, protein structure prediction is not fully equivalent to protein folding.

Yeah, this is what I mean by "a shitty answer fast" - structure prediction isn't a canonical answer, but it's a good enough approximation for good enough decision-making to make a bunch of stuff viable that wouldn't be otherwise. I agree with you, though - they're two different answers. I've done a bunch of work in the metagenomics space, and you very quickly get outside areas where Alphafold can really help, because…

> this is what I mean by "a shitty answer fast" - structure prediction isn't a canonical answer

A proper protein structural model is an all-atom representation of the macromolecule at its global minimum energy conformation, and the expected end result of the folding process; both are equivalent and thus equally canonical. The “fast” part, i.e., the decrease in computational time comes mostly from the heuristics used for conformational space exploration. Structure prediction skips most of the folding pathway/energy funnel, but ends up at the same point as a completed folding simulation.

> At that point, an actual solution for protein folding that doesn't require a supercomputer would make a difference.

Or more representative sequences and enough variants by additional metagenomic surveys, for example. Of course, this might not be easily achievable.

Re: Chemistry Nobel: Computational protein design and protein structure prediction

#284

Post by PaulDavisThe1st »

concordDance wrote: post #265
PaulDavisThe1st wrote: post #238

Earlier quoted context omitted.

That's one metric, that only reflects Amazon's function as an income generator. I view businesses through other metrics as well, including their impact on society in a variety of different ways. From some of those perspectives, it is not clear to me that Amazon (where I was the 2nd employee) is a net benefit.

This Amazon the company specifically or online shopping in general? E.g. if Amazon hadn't been made and some other online retailer had dominated (or even if there had been many!)

Both.

Re: Chemistry Nobel: Computational protein design and protein structure prediction

#285

Post by throwaway48476 »

BurningFrog wrote: post #189
throwaway48476 wrote: post #81

Earlier quoted context omitted.

Perhaps it's time the prize goes to the discovery itself instead of a person.

How do you give $1M to a discovery?

$1M ain't what it used to be.

Re: Chemistry Nobel: Computational protein design and protein structure prediction

#286

Post by jerjerjer »

mk_stjames wrote: post #218
ThePhysicist wrote: post #67

Demis Hasabis has a really interesting and unusual CV for a nobel laureate [1], he started his career in AI game programming (he worked e.g. on Popoulous II, Syndicate, Theme Park for Bullfrog, and later for Lionhead Studios on Black & White) before doing a PhD in neuroscience, becoming an entrepreneur and starting DeepMind. I would say this is a refreshing and highly uncommon pick for a nobel prize, really cool to s…

I'm always interested in hearing about these people who go and get a PhD in an unrelated field to their original studies, often years after leaving university and working in an industry. Here it says Hasabis did an undergraduate degree in a computer science program, and them spent a decade working on computer games at studios, and then somehow just rocked up to a university and asked to do a PhD in neuroscience. I fe…

PhD is a thankless, low paid position with insane hours and zero guaranteed return. Outside of a few elite programs and universities getting into PhD program is fairly easy - they take anyone qualified.

Re: Chemistry Nobel: Computational protein design and protein structure prediction

#287

Post by roughly »

flobosg wrote: post #282
roughly wrote: post #264

Earlier quoted context omitted.

Yeah, this is what I mean by "a shitty answer fast" - structure prediction isn't a canonical answer, but it's a good enough approximation for good enough decision-making to make a bunch of stuff viable that wouldn't be otherwise. I agree with you, though - they're two different answers. I've done a bunch of work in the metagenomics space, and you very quickly get outside areas where Alphafold can really help, because…

> this is what I mean by "a shitty answer fast" - structure prediction isn't a canonical answer A proper protein structural model is an all-atom representation of the macromolecule at its global minimum energy conformation, and the expected end result of the folding process; both are equivalent and thus equally canonical. The “fast” part, i.e., the decrease in computational time comes mostly from the heuristics used…

> ends up at the same point as a completed folding simulation.

Well, that's the hope, at least.

> Or more representative sequences and enough variants by additional metagenomic surveys, for example. Of course, this might not be easily achievable.

For sure, but for ostensibly profit-generating enterprises, it's pretty much out of the picture.

I think the reason an actual computational solution for folding is interesting is that the existing set of experimentally verified protein structures are for proteins we could isolate and crystalize (which is also the training set for AlphaFold, so that's pretty much the area its predictions are strongest, and even within that, it's only catching certain conformations of the proteins) - even if you can get a large set of metagenomic surveys and a large sample of protein sequences, the limitations on the methods for experimentally verifying the protein structure means we're restricted to a certain section of the protein landscape. A general purpose computationally tractable method for simulating protein folding under various conditions could be a solution for those cases where we can't actually physically "observe" the structure directly.

Re: Chemistry Nobel: Computational protein design and protein structure prediction

#289

Post by dekhn »

flobosg wrote: post #282
roughly wrote: post #264

Earlier quoted context omitted.

Yeah, this is what I mean by "a shitty answer fast" - structure prediction isn't a canonical answer, but it's a good enough approximation for good enough decision-making to make a bunch of stuff viable that wouldn't be otherwise. I agree with you, though - they're two different answers. I've done a bunch of work in the metagenomics space, and you very quickly get outside areas where Alphafold can really help, because…

> this is what I mean by "a shitty answer fast" - structure prediction isn't a canonical answer A proper protein structural model is an all-atom representation of the macromolecule at its global minimum energy conformation, and the expected end result of the folding process; both are equivalent and thus equally canonical. The “fast” part, i.e., the decrease in computational time comes mostly from the heuristics used…

Most proteins don't fold to their global energy minimum- they fold to a collection of kinetically accessible states. Many proteins fail to reach the global minimum because of intermediate barriers from states that are easily reached from the unfolded state.

Attempting to predict structures using mechanism that simulate the physical folding process waste immense amount of energy and time sampling very uninteresting areas of space.

You don't want to use a supercomputer to simulate folding; it can be done with a large collection of embarassingly parallel machines much more cheaply and effectively. I proposed a number of approaches on supercomputers and was repeatedly told no because the codes didn't scale to the full supercomputer, and supercomputers are designed and built for codes that scale really well on non-embarassingly parallel problems. This is the reason I left academia for google- to use their idle cycles to simulate folding (and do protein design, which also works best using embarassingly parallel processing).

As far as I can tell, only extremely small and simple proteins (like ribonuclease) fold to somewhere close to their global energy minimum.

Re: Chemistry Nobel: Computational protein design and protein structure prediction

#290

Post by dekhn »

flobosg wrote: post #222
nextos wrote: post #221

Earlier quoted context omitted.

AlphaFold is excellent engineering, but I struggle calling this a breakthrough in science. Take T cell receptor (TCR) proteins, which are produced pseudo-randomly by somatic recombination, yielding an enormous diversity. AlphaFold's predictions for those are not useful. A breakthrough in folding would have produced rules that are universal. What was produced instead is a really good regressor in the space of proteins…

> A breakthrough in folding would have produced folding rules that are universal. Protein folding ≠ protein structure prediction > I think those who invented pairwise and multiple alignment dynamic programming algorithms deserved some recognition I would add BLAST as well but that ship has sailed, I’m afraid.

The value in BLAST wasn't in its (very fast) alignment implementation but in the scoring function, which produced calibrated E-values that could be used directly to decide whether matches were significant or not. As a postdoc I did an extremely careful comparison of E-values to true, known similarities, and the E-values were spot on. Apparently, NIH ran a ton of evolution simulations to calibrate those parameters.

For the curious, BLAST is very much like pairwise alignment but uses an index to speed up by avoiding attempting to align poorly scoring regions.

Post reply on HN