siver_john

I don't know how you would classify this for replace vs augment. However Neural Network force fields which replace traditional forcefields in simulations are and have been an active area of work. There are plenty of examples but I'd say most notable are Deep-MD by the Weinan E lab of Princeton (they even have a docker file on NGC), the work of Behler and derivatives, and probably more towards the direct replacement, TensorMol which I've seen some recent movement on.

Using Neural Networks to solve differential equations have been a hot topic under Physics Informed Neural Networks at the last few NVIDIA GTCs. And arguably my work can also fall under a similar umbrella as well. So your lab was definitely before the curve.

In my time as a PhD student (granted still there currently just at the finishing things up stage) I was actually part of founding an organization for students to basically do this. We invited speakers from different fields and had them talk about "alternative" career paths considering the current system basically assumes you want to become a professor. We also included a workshop every year where we would build in career skills and have professionals come and speak. You learn very quickly that there is an incredible range of activities one can do with a PhD.

That's a cool way I had never thought about it being used. Just for my own curiosity do you know if that is a common use case in the protein solving field? (I'd imagine it would also be useful in NMR experiments for getting initial point labels).

My memory is a little rusty, but I believe yes. If I remember correctly simulation of the virus helped medically in a few ways. Specifically I think it gave insight into the mRNA vaccines and what sequences to use to make them effective (by basically making a slightly worse spike protein). I am sure it helped in drug discovery as I know our lab used simulations to suggest some potential drug pockets. There were some really good talks about it at NVIDIA's GTC last year or so (maybe more at this recent GTC but I had too much going on personally to watch the VODs).

As an insider to both fields (and specifically their juncture to molecular dynamics).

What you are discussing is of course being studied but the problem is that it is a lot more computationally expensive. We do have simulations of simple (read very small) proteins folding and unfolding but for larger ones the computational time to watch them fold can be gigantic, if not impossible due to the fact that proteins often fold as they are being made. Which means including a much larger process into a folding one which just further stresses computational resources.

This computational problem is so enormous that a company at the cutting edge of research D.E. Shaw built a specialized computer solely for simulating proteins. Also most of the software used for this until recently had abandoned multi-GPU paralellism because it didn't scale well. The pandemic caused the need to simulate the virus on the entirety of Summit and introduced some work back into that route but it is still specialized (and wouldn't help for systems below a certain size anyways).

Also my previous points have been for atomic models (e.g. we treat everything more like newtonian particles and ignore quantum effects) some things definitely need more resolution and at that level you are lucky to see protein fluctuations let alone folding.

I always enjoyed forums that had "Newbie Corners" where newbies could go to introduce themselves if they wished, otherwise I rarely saw the newbie welcoming type of interaction outside of that corner. (Unless it was obvious the newbie had run afoul of one of the forum's rules but that's on the newbie at that point.)

If I'm reading your comment correctly, then fingering has the same connotation in the U.S. as well, it is also synonymous with third base. (Also have never heard the term either.)

I mean nothing is stopping you from skipping validation with pre-alphafold techniques and say for drug discovery to already do drug screening using the predicted structure. It's just the drug screening software is already error prone so you are still going to have to do some validation. However having an idea on a potential structure means that you can do other techniques that are simpler to validate it that are less expensive/time consuming (I'm thinking of things similar to FRET).

Another idea is these may come into play for anti-verification, so if you are drug screening against a known structure. You could potentially use these more flawed structures of proteins you don't want to target but may be similar, and try to reduce the drug's efficacy at binding them. Or something to that effect. All of that is fun ideas that are currently being explored in that space but we'll see where it takes us.

Looking at Google Scholar it appears the related paper is here: https://iopscience.iop.org/article/10.1088/1361-6463/ac006f/...

There also appears to be a patent but I don't trust my googlefu enough to try and search for that.

I am not an MIT researcher however my university dropped Elsevier at some point and I think generally the experience and sentiment is similar across universities and fields from the admittedly small sample size I have.

It makes things more annoying on some levels, but between the two sides of university library versus Elsevier the large majority of scientists agree with Elsevier being wrong so the annoyance is generally acceptable. Ideally this entire system changes for the better but we can make it work in the short term. Not to mention that even before this we often ran into journals we didn't have access to, so some of these annoyances would have existed regardless.

I had this problem a few months ago, but because Sci-Hub doesn't have supporting information currently. Yes, I understand normally that is free anyways on journals but for whatever reason this journal (owned by Elsevier ofc) did not have its SI available without a subscription and my institution did not have access to this journal. So no, just use sci-hub is not a valid alternative as it stands today and because of its legal jeopardy means we probably shouldn't become reliant on it.

Not to be that guy, but isn't BSD and it's derivatives still Unix derived so parent's comment still holds? (Though to the parent comment I'd argue that OSX isn't really a sold OS either because I can't buy just it)

Yes, ludicrously so, platinum is one of the most expensive metals.

Edit: For actual values (per gram), Gallium ~0.50 USD, Platinum ~33.22 USD

Just for the record enzyme does not mean organism. Enzymes are non-self replicating so the "grey goo" scenario isn't a problem until we introduce them into organisms but then the energy they derive from the plastic must be greater than the rest of the machinery used to turn it into energy/nutrients. And I would be skeptical of that any time soon.

It is often a model organism in plant studies, similar to mice in drug discovery. Mostly because it has a quick growth time, is fairly hardy, and just the fact it has been so well studied means that a lot is known about it.

I am not a plant biologist but have talked to some and I've often seen it used in their work and this was their given reasons.

  Location: Burlington, VT
  Remote: Yes
  Willing to relocate: Yes
  Technologies: Python, CUDA, Tensorflow, Schrödinger, Amber
  Resume/CV: https://drive.google.com/file/d/1m6mybONmVFO1RFHtxhIpKd9_lUcaO57T/view?usp=sharing
  Email: jferre14@vols.utk.edu

I am a graduate student planning on defending soon with a wide background in STEM generally though my field can be most charitably characterized as Computational Biophysics. I have a strong background in molecular dynamics and applying AI to that domain. Feel free to reach out if you want to know more.

Edit: Formatting

As long as you can clearly define your starting axioms you can mathematically prove or disprove anything. The physical example would be to just say, suppose a universe with the same properties of ours exist but where the gravitational constant is twice as large prove whether X can occur in that universe.

As far as how randomness is defined I believe that might be field dependent but this is me getting out of my depth (I've only taken a few courses on combinatorics).

Just to add my two cents as someone who recently switched to Colemak-DH after getting a moonlander (I used a das ultimate for years though), but I am currently typing this on my laptop using QWERTY.

I feel like switching between the two isn't always 100% painless with a small lag time especially if I've been away from one or the other a few days while using the other one predominantly. But generally it's not too bad. The worst of it being typing on my laptop, messaging friends late at night and having a sudden forgetfulness which is easily overcome by just looking at the keyboard for a moment while typing till my hands remember.

Considering I often come to things like Show HN/Ask HN to see what exists in the world, yes. I say it is not only appropriate but within the spirit of fostering good conversation and interest in the community. (Obviously only as such competition remains civil.)

The graph on here seems similar to what I've noticed. My lab mainly uses Tensorflow mainly driven by my knowledge. And the only reason why I learned Tensorflow initially was that PyTorch was just starting when I was choosing a framework and the documentation wasn't as established. However, recently I recommended a student who was asking me which framework to probably use PyTorch due to the ease of implementation comparatively.

I can't imagine the infrastructure to do it is trivial, but I would say the larger burden is network effect. Bandcamp has existed for a long time and I know of indie record labels who use it as their default distribution. I am aware there are indie alternatives (some of which provided in the comments to my first comment). Also disruption of service often results in loss, would someone who is no longer focused on music move there stuff over if things changed dramatically? Would people download in time, etc?

Obviously this is all speculation, bandcamp could continue on as it has been for the conceivable future, but I am less pleased about that future than I was before I saw this news.

I do through my purchases and my general encouragement of traffic to the site, if I wasn't a broke graduate student and had the funds to invest in them, without a doubt in my mind I would.

However, even if I am unable to invest in them I see nothing wrong about my expressing discomfort over someone else buying them. I have seen nothing as well that they needed cash to continue operations.

You are a god send. If for no other reason than they have an album I had been looking to purchase I could find nowhere else.

This feels like a major blow, as someone who just semi recently (probably around pandemic start) started getting into purchasing flac music from indie artists, Bandcamp was a great source of music. I understand nothing will change in the short term, but long term I am very concerned. Especially as streaming becomes more dominant and companies are less willing to provide flac based music and physical discs (where I can rip my own) continue to disappear.

This feels like a potential last step of true music ownership and that makes me incredibly sad.

That being said if anyone knows of any place to buy flacs of music with great selection would love to know (especially for Japanese music which I generally have to import, thankfully they love CDs).

Thank you for sharing, I had been looking for this after seeing it on here the other day but I couldn't remember the name or anything at all.

Just as a point, your argument is a false dichotomy. The only two choices you allow are a ruthless capitalistic society or a communist society. The poster could have allowed for any colorful regions of societies in between.

I would argue most societies throughout history have treated people with mental illness less than kindly. But if wish to treat them better we must stop thinking of people as factors of productivity, and therefore their worth is only as useful as the amount they add to the bottomline. (Though I'd presume, albeit with little evidence currently, that society would be benefited in probably cost and happiness if we were to take on the task of providing adequate mental care to all.)

Just imagining the mathematics SWAT team coming through the door and swatting the marker out of one of my physics profs' hands when they went from ydy/dx=x -> ydy=xdx as if they were simply re-arranging a fraction, made me chuckle.

Anyone who is going to be buying PoS from Square at a rate greater than just the phone attachment reader, this won't be sufficient for. Using a phone for a business you personally run is fine, using a phone as the central point of your business for even a relatively small fast casual is going to be a nightmare.

And for those even on Square Apple's ipads are still the preferred choice as far as I can tell. So there isn't much of a benefit other than fees.

For anything larger than a small fast casual you quickly run into greater integration needs with things like KDS (of which a few do use tablets, Toast, Fresh, etc, and even then I think ipads are preferred. I know Square integrates with a few of those as well as NorthStar so, I don't think they may have much to fear other than at the low end mom and pop stores.

>Actually, a whole lot of recombinant approaches have moved to human trials (along with viral vector approaches).

I should have been more clear here, I meant specifically trials of what I presume could have been attempted if the vaccine was expressed as protein instead of mRNA. But thanks for the clarification. And the mention of viral vectors which are an area I was less aware of till recently but it has been cool to see them in more places.