Skip to main content
57 votes
Accepted

DeepMind just announced a breakthrough in protein folding, what are the consequences?

Imagine if given an amino acid sequence, you could quickly calculate what the shape of the corresponding protein would be. You would be able to predict what effect a mutation would have on the shape ...
Nike Dattani - No Free Time's user avatar
23 votes
Accepted

What is the current status of machine learning applied to materials or molecular systems?

Here is state of the art research: Smith J.S. et al, Approaching coupled cluster accuracy with a general-purpose neural network potential through transfer learning, July 2019 Nat. Commun. 2019, 10 (1)...
Peter Morgan's user avatar
  • 1,483
17 votes

What is the state of the art in terms of local atomic environment descriptors for machine learning?

If you are familiar with the Behler-Parrinello symmetry functions implemented in AMP, you may be interested in seeing how they compare to other atom-centered representations in terms of speed and ...
Stephen Xie's user avatar
17 votes

Machine learning interatomic potentials for molecular dynamic simulations: are they good?

Before talking about the pros and cons of ML-potentials, there is a huge conceptual difference between empirical- and ML-potential that needs to be clarified. In empirical potentials, one uses data ...
Saleh's user avatar
  • 271
17 votes
Accepted

Machine learning interatomic potentials for molecular dynamic simulations: are they good?

(Expanding my comment into an answer) When ML-based forcefields are compared to classical forcefield directly, I think we miss the most important points. ML-based models have several advantages: They ...
Greg's user avatar
  • 1,807
17 votes
Accepted

Does DeepMind's new protein folding software (AlphaFold) also work well for metalloproteins (proteins with metal cofactors)?

It's a great question! Some of my answer will be taken from my answer to your question on the AI stack exchange, but cross-site questions are allowed and your question here is slightly different so my ...
Nike Dattani - No Free Time's user avatar
16 votes
Accepted

What does machine learning learn about DFT?

Total energies and HOMO-LUMO gaps are very different quantities, and naturally necessitate very different neural network designs (including the choice of descriptors and architectures) in order to ...
wzkchem5's user avatar
  • 9,670
15 votes

Benchmark Timings of Machine Learning Potentials vs Molecular Mechanics Force Fields

We performed some timing benchmarks as part of our recent paper, albeit not on molecular dynamics: "Assessing conformer energies using electronic structure and machine learning methods" Int ...
Geoff Hutchison's user avatar
15 votes
Accepted

General Techniques for Smart Sampling in Matter Machine Learning?

This is not an exhaustive answer. This is an evolving research area in terms of applying ML to dataset generation. I am most familiar with the use case for constructing atomistic potential energy ...
jheindel's user avatar
  • 3,584
15 votes
Accepted

Artificial intelligence is a hot topic, but should I pursue it if I'm interested in Matter Modeling?

"I am a student now and should decide my major to research soon. AI/ML is a very hot research field and I am very interested in it, but I have some doubts before I study AI/ML." AI is ...
Nike Dattani - No Free Time's user avatar
14 votes

What is the current status of machine learning applied to materials or molecular systems?

Machine learning (ML) is rapidly gaining its popularity in the field of materials science due to its exceptional ability to learn from data to guide experimentalists, thus reducing traditional trial ...
Achintha Ihalage's user avatar
14 votes

How to start a Machine Learning project for chemical properties prediction?

Admittedly, there are tons of materials on the chemistry + machine learning topics. Let me give one: An introductory text I find useful is Machine Learning in Chemistry from Janet and Kulik in ACS in ...
Greg's user avatar
  • 1,807
14 votes
Accepted

What are the advantages of (semi)-empirical force fields over Machine Learning Potentials?

The answer to this question is inevitably going to be opinionated. My opinion is that there are still very good reasons to explore the development of force fields while also pursuing better ML ...
jheindel's user avatar
  • 3,584
13 votes
Accepted

Deep Neural Networks: Are they able to provide insights for the many-electron problem or DFT?

"However, it is notorious due to the exponential wall" That is completely true, though there's indeed some methods such as FCIQMC, SHCI, and DMRG that try to mitigate this: How to overcome ...
Nike Dattani - No Free Time's user avatar
12 votes
Accepted

What are some examples of active learning methods used in atomistic machine learning?

Inverse designing of materials with known target properties is of great importance (to reduce time, labour, financial etc. costs) than the traditional way of materials design which is guided by human ...
Achintha Ihalage's user avatar
12 votes
Accepted

Getting interpretable chemical information from hashed molecular fingerprints

Using RDKit, it's fairly easy to depict bits on example molecules - it's even an example in the documentation "Generating Images of Fingerprint Bits") ...
Geoff Hutchison's user avatar
11 votes

If total energies differ across different software, how do I decide which software to use?

"how do I decide on which algorithm/software to use if the total energies are different across algorithms/software?" You can use any algorithm or software to calculate your total energies, ...
Nike Dattani - No Free Time's user avatar
10 votes

What does machine learning learn about DFT?

It could just be that the features you are using are well suited for describing total energies in these sorts of systems, but not in describing the differences of eigenvalues. When people do ML to try ...
AGS's user avatar
  • 1,141
9 votes
Accepted

What are some available software packages for automated finding of local and absolute minima on PES?

The Atomic Simulation Environment has two nice implementations of global optimization algorithms. The first is a basin hopping algorithm from a 1997 paper by Wales and Doye in J. Phys. Chem. A. The ...
Andrew Rosen's user avatar
  • 7,391
9 votes

Can Machine Learning lead to the more accurate theories and methods for matter modeling?

It is certainly possible to develop ML models that yield more accurate results than would be possible without ML. One route to do this is through so-called "Δ-learning" where you use ML to ...
Andrew Rosen's user avatar
  • 7,391
9 votes

Best resources for someone going from experimental to computational chemistry

There are advantages for someone who has experience with both experiments and simulations. You don't mention your specific experimental expertise, but looking for topics (or postdocs) that bridge ...
Geoff Hutchison's user avatar
9 votes
Accepted

Course on machine learning on application in material science

Machine Learning for Physicists If you wish for a general introduction to a number of ML concepts and tools useful in physics (not entirely specific to "material science"), and taught for ...
Hebo's user avatar
  • 1,675
8 votes
Accepted

Is there a database where one can find the Electron Density data of materials?

I recently made charge densities available for the MOFs and coordination polymers of the Quantum MOF (QMOF) Database. Please read the GitHub page for details on how to access the charge densities. ...
Andrew Rosen's user avatar
  • 7,391
8 votes

What is the best way to measure similarity between molecules of the same formula?

tldr; The most common approach is to use fingerprints and compute the Tanimoto similarity There are a variety of ways to compute "molecular similarity" but the most common approach is to ...
Geoff Hutchison's user avatar
8 votes

Artificial intelligence is a hot topic, but should I pursue it if I'm interested in Matter Modeling?

I did my master's research project in predicting absorption spectra using machine learning (with fingerprints). So, my research is somewhat in the field of material modelling you can say. I want to ...
S R Maiti's user avatar
  • 7,011
8 votes

How can I build the wheels necessary for a quick installation of PySCFad on a "compute node"?

I finally got it working! Explanation of the error The above procedure installed the "main" branch of "pyscf_properties" rather than the "ad" branch. It seems that due ...
Nike Dattani - No Free Time's user avatar
8 votes

How a beginner should start his studies in ML for chemistry application?

You state your background is chemistry, and have aspiration for Python as a programming language aiming for ML without background in computer science. Do you have some experience in programming at all?...
Buttonwood's user avatar
  • 1,579
7 votes

Can Machine Learning lead to the more accurate theories and methods for matter modeling?

Within Monte Carlo (MC) methods, there are a few areas of active research in this regard: Training ML models to identify phase transitions: In practice, it challenging to identify phase transitions ...
taciteloquence's user avatar
7 votes

DeepMind just announced a breakthrough in protein folding, what are the consequences?

Keep in mind that many if not most proteins have multiple quasi-stable conformations, so their 3D structure is not actually a single conformation but rather a Markov matrix of conformations, with ...
Eric Minch's user avatar
7 votes

DeepMind just announced a breakthrough in protein folding, what are the consequences?

Probably one of the important applications is Computer Aided Drug Discovery (CADD). If the protein structure could be accurately predicted, one could design protein-ligand docking on the binding ...
Paulie Bao's user avatar
  • 3,973

Only top scored, non community-wiki answers of a minimum length are eligible