Talk:Molecular dynamics/Archive 1

Archive 1

Article is too centered on biomolecules

OK, so I keep adding things about biomolecules because that's what I model. But I would ask people working in other fields with MD (i.e. metals) to add examples of applications. Jorgenumata 21:42, 22 October 2006 (UTC)

Hi Jorgenumata. I liked all the additions you have made. I myself have worked in metals on the physics side, so I can add some things. But actually, for the overall structure, I think it would be useful to discuss some of the ways that MD is used in different fields on the talk page so that we can incorporate the different uses of MD more fully. Most people are focused on their on field, and don't see outside it (me included), so it is kind of hard to incorporate all the different things into one article.

The merge discussion with MM is an example. Metals cannot be modelled by force fields. They are completely incompatible with the force field view of molecules, and from reading the MM page, MM only uses force fields. So it seems impossible to merge the articles. Anyway, enoough of my blabbing. The things that I am interested in knowing are what/how MD is used in different fields.

Fluid/aerodynamics: Do fluid dynamics / aerodynamics people use the billiard-ball approximation of atoms to model fluids, and are these methods considered molecular dynamics if they consider each atom individually?

Metals: We don't use force fields. Generally, TBSMA or equivalent is the least accurate/most empirical that can be used. Tight-binding and ab initio are far more common. Anyway, I can add a lot of metals things myself, so don't need much input here.

Semiconductors: Again, force fields are not used. However, requires very different approximations to metal simulations. Again, I can add stuff to the article on these.

Chemists: What exactly do chemists study? Are the chemical reactions always studied within a solvent, or are they studied on isolated chemical compounds? Is it always important to correctly model the temperature and pressure, or are these secondary considerations in chemical MD simulations?

Bio: Similar to chemistry, are temperature and pressure important? Do proteins get modelled within the context of a solvent, or are conformations studied on the basis of a single protein molecule with no external interactions?

So, I'll try to add some metals and semiconductor things to the article, but I don't want to break the chem/bio side. Will come back in maybe a week and try to do some additions. Mattopia 09:09, 21 November 2006 (UTC)

Merging of article Molecular Mechanics

I think the article on molecular mechanics is redundant and should be merged into this one, as well as some parts into force field (chemistry). It is completely equivalent, even from the content. The term "molecular dynamics" is nowadays more common than "molecular mechanics", except for the fixed expression QM/MM (quantum mechanics/molecular mechanics). What do you think? Jorgenumata 21:22, 9 October 2006 (UTC)

I disagree. The term is widely used in chemistry and molecular biology where the main thrust is to predict the equilibrium geometry. This is not dynamics. --Bduke 21:01, 10 October 2006 (UTC)

To expand on this, if you look at "Computational Chemistry" by David Young, "Essentials of Computational Chemistry" by Christopher Cramer and "Molecular Modelling" by Andrew Leach, you will find the two topics dealt with in different chapters. These are three of the key texts in computational chemistry. --Bduke 06:00, 11 October 2006 (UTC)

I agree with Bduke. MM is certainly much more narrowly focussed, than MD, on molecular geometry, especially to chemists. I do not agree that it should be merged with MD. vgsbox 23:58, 10 Oct. 2006

Also see Talk:Molecular_mechanics for more opinions... Itub 01:14, 15 October 2006 (UTC)

Force Fields

I added some stuff from a physics point of view, but I do struggle a bit with some of the terms used by chemists. In particular, what exactly is a force field? My supervisor used to say that what the protein-folders called a 'force field' was the same as what physicists called 'tight binding', but on reading the articles on force fields, that doesn't seem correct. Can anyone help ? Mattopia 07:32, 28 October 2005 (UTC)

Please see the distinction between force field (physics) and force field (chemistry). Karol 12:20, 28 October 2005 (UTC)

A succinct way to put it is that the force field specifies the form of the Hamiltonian for a classical system. As the force field (chemistry) article describes in some detail, most classical forcefields assume the dominant terms of their Hamiltonian to be some form of Van Der Waals force and some sort of coulombic electrostatic force and then one must come up with a set of VdW coefficients and partial charges deemed to be accurate enough for the problems of interest. There's some other spinach too but that's the main ideaRobotkid 12:06, 2 August 2006 (UTC)

• I suggest moving part "Potentials in MD simulations" to Force field (chemistry) unless anyone objects. My very best wishes (talk) 04:38, 3 March 2012 (UTC)

You've replied to a section from six years ago, I suggest it may be best to reformulate the argument for a split in a new section perhaps. IRWolfie- (talk) 10:33, 3 March 2012 (UTC)

Definition of MD

This article is misleading! The term "molecular dynamics" means simply the motion of molecules, such as translation, rotation, vibration, etc. Experimentally molecular dynamics can be probed by spectroscopy (Infra-Red, Microwave, Raman, etc.) and scattering etc. Molecular dynamics simulation is a computer simulation method that has been developed to understand the measured spectra and mainly to gain a deeper understanding of the motion of molecules. It aims to mimic the motion of molecules and atoms, and it aims to be consistent with spectroscopy. In theoretical circles, however, it is true that molecular dynamics refers usually to the simulation method, but that is not necessarily true among experimentalists. Galileo fan 10:29, 7 December 2005 (UTC)

I disagree - taken individually both the words "molecular" and "dynamics" are incredibly vague - even among experimentalists the use of such imprecise language becomes a problem because of the huge disparity in length-scales and time-scales between different experimental techniques (in NMR spectroscopy alone, "dynamics" can be measured from the picoseconds to the millisecond range). In terms of simulation, however, it refers to a very specific method of sampling via the integration of equations of motion which is not necessarily motivated by a need to reproduce spectroscopic data. In principle, any equilibrium property can be measured with molecular dynamics as a sampling technique (for example diffusion coefficients, osmotic/activity coefficients, dielectric constants). Robotkid 12:06, 2 August 2006 (UTC)

Molecular dynamics is quite a generic term. I suggest to split this article. Galileo fan is right. What it is in the current article must be classified as "Molecular Dynamics Simulation Methods, Techniques and Algorithms" Please let me know your ideas. Otherwise I will split the article.

Also subsection "Types of MD Systems" is not mathching with the sub-sections. "Force fields" or "inter-atomic potentials" is not a type of MD systems, simply they are the part of the technique.

--mcyp 09:00, 10 August 2006 (UTC)

Statistical Ensembles

I have edited some part. Pointing out integrators and their use in the corresponding Statistical Ensembles. Also, added senteces that corresponding Mathematical field on "Molecular Dynamics". There was an error that MD gives time dependent properties. That is a misconception. Equilibrium MD only gives static properties (time-independent or steady-state) where it is also based on Ergodic theory. Cheers

--mcyp 12:28, 2 August 2006 (UTC)

Hi Msuzen, I was wondering if you could move some of your additions into a sub-section on statistical mechanics and equilibrium/non-equilibrium systems in MD. When I first read your changes it took a little while to understand what you were saying, and I think it could do with more explanation. Unfortunately, I don't think I have the necessary knowledge to do it myself. Mattopia 15:14, 2 August 2006 (UTC)

Hello Mattopia, If you refer to having a new sub-section, it is a good idea. May we can add "Ensembles and Integrator Algorithms" and "Non-Equilibrium MD" sections? Any ideas?

--mcyp 08:43, 10 August 2006 (UTC)

it is the Gibbs hypothesis that time average equals the ensemble average

Torsion angle dynamics

I guess it is important to say something about torsion angle dynamics applied in DYANA and IVM module of XPLOR. Biophys 17:45, 31 October 2006 (UTC)

General to-dos

Posting here for wider reference.

• Lead needs major work
• General organization is awkward; algorithms section is ridiculously stubby
• Rather little discussion of the implicit/explicit solvent decision
• Barely anything on constraints??
• Not terribly important, but grand canonical ensemble is missing
• Problems of current force fields not covered/glossed over; should get a brief mention
• Numerical issues not covered or barely covered - energy drift not even a brief mention? integration schemes? (only verlet gets even a link) and timesteps barely even get a sentence
• PBC issues should be covered, though maybe best split into a subarticle (which is totally missing at the moment)

Opabinia regalis 04:10, 27 January 2007 (UTC)

Integrators

The integrator section needs to be redone. It contains several errors.

First of all, the Nose-Hoover temperature control is not an integrator; it's just a thermostat which gives some additional differential equations which control the system temperature. One still needs some algorithm to solve the equations. Same goes for Parinello-Rahman pressure control. These do not belong to the integrator section. In fact I'm going to remove them after this post.

Secondly, it's wrong to say that Verlet integrator is somehow microcanonical. There's no reason why one couldn't use some temperature/pressure control with it. Ossi 13:20, 8 May 2007 (UTC)

Large animation

Isn't the >600K animation a bit big to be in an article? Remember, some people are still on dial-up. 210.50.115.191 02:48, 24 September 2007 (UTC)

Ab-Initio Section Highly Illegible

This section needs work on general grammar and style and readability. 23:37, 4 October 2007 (UTC)

Add a tag? Pedro Gonnet 08:26, 5 October 2007 (UTC)

I just reworked and reorganized the ab-initio section. Still, we need an expert on this to tend to it. Jorgenumata 09:06, 8 October 2007 (UTC)

Databases of trajectories relevant to topic

These links to databases of molecular trajectories have been deleted twice. I think they are perfectly relevant to the topic of molecular dynamics. Jorgenumata 16:12, 31 October 2007 (UTC)

• Dynameomics.org: Home page of the Dynameomics project and database of molecular dynamics simulations for several hundred folds
• BioSimGrid, a database for biomolecular simulations
• MolMovDB: A database of molecular motions

I'm the one who removed them. To my understanding, this article is more about molecular dynamics in general, which these links are not. There are more specific pages, such as Protein folding, Molecular mechanics, Molecular Modelling Protein structure prediction, Software for molecular mechanics modeling, List of protein structure prediction software, etc... where these links would be more at home.

Cheers, pedro gonnet - talk - 01.11.2007 08:10

Hola Pedro. Do you mean that this page should only be about the technical aspects of simulation, and not about the results, application and analysis? I think these links are useful for the readers to understand what you can do with MD. All the wikipedia entries you mentioned deal mainly with static concepts of proteins. These external links offer dynamic trajectories, which are the product of molecular dynamics simulations. Jorgenumata 13:18, 4 November 2007 (UTC)

Sorry for answering kind of late -- I'm out sick... I agree that Molecular trajectories are important enough and I guess the best thing to do would be to start a specific article on that topic and link it in the "See also" section of this and related articles. That way we resolve the problem of where things go and have all the relevant links on this topic. What do you think? Cheers, pedro gonnet - talk - 06.11.2007 11:01

Hola. Honestly I don't think molecular trajectories deserve a separate page. They are the sole product and objective of a molecular dynamics simulation, which is then to be analyzed. 160.45.24.162 (talk) 09:40, 20 November 2007 (UTC)

Major software for MD simulations

The list of software for MD simulations (and the section "Related software") is out of control. I suggest merging the whole thing into the article List of software for molecular mechanics modeling and leaving either a short paragraph or just the link in "See also". If nobody opposes this move, I will follow through.

Cheers, pedrito - talk - 06:29 16.09.2009

A few comments

I see two shortcomings in this article. First, the subject is not properly defined. Simply speaking, the method of MD means solving classical mechanics equations for a system of molecules and atoms to determine their trajectories. No, it is not based on statistical mechanics, but it can use statistical mechanics to interpret the results of calculations. Second, the article includes descriptions of force fields, which are basically the same as in molecular mechanics an molecular modeling in general. Those should probably be moved to "force fields" article. If there are no objections, I will make some changes later.Biophys (talk) 19:00, 31 January 2011 (UTC)

!!!PLEASE !!! do so. I was about to post a comment with essentially the same wording. I am trying to edit the article about Charles Coulson. The opening sentences link to articles that are disastrously wrong. I have tried the tactic of inserting a new lede, then heading the previous lede "An alternative approach" to sidestep the bottomless pit of rewriting the rest. The Discussion page of the theoretical chemistry article shows that it was put together mostly by people (?high school children?) identified by IP number, and a pseudonymous editor whose user page list articles he considers his product, that include one on prediction of protein structure that does not mention any of the theoretical work that uses methods to solve polynomial equations, or the University of Washington Foldit project. One contributor reports asking a graduate student what theoretical chemistry is! I am conducting an experiment at Talk:Applied_mathematics#Suggestion for revised lede, draft 3. Michael P. Barnett (talk) 15:39, 4 February 2011 (UTC)

Too much effort. The problems are difficult to count. This article should be moved/renamed to Molecular dynamics simulations. "Molecular dynamics" per se can be studied by experimental methods, like NMR spectroscopy, and that is what should be described here. Biophys (talk) 01:22, 15 June 2011 (UTC)

I fixed the intro in first approximation. Biophys (talk) 15:50, 15 June 2011 (UTC)

Molecular dynamics is not strictly only a classic treatment. IRWolfie- (talk) 15:14, 27 November 2011 (UTC)

Maybe a split into four articles would make sense: A split between Ab Initio and related methods, Mixed methods, classical methods and potentials/force fields. IRWolfie- (talk) 15:20, 27 November 2011 (UTC)

Broken external links

The three introductory .pdf articles appear to be no longer available. — Preceding unsigned comment added by Memcbr (talk • contribs) 20:56, 7 March 2011 (UTC)

Lede is too focused on classical physics

The lede makes it sounds like Molecular Dynamics is only done through classical physics. Can it be made more clear in the lede what applies only to classical conditions? Also a mention of dynamics where quantum physics is required should be made in the lede. IRWolfie- (talk) 22:21, 25 November 2011 (UTC)

If you want to add something about quantum mechanics or a "hybrid approach", please do. My very best wishes (talk) 04:32, 3 March 2012 (UTC)

Section is unclear

This section Molecular_dynamics#Potentials_in_MD_simulations appears to imply that the Born-Oppenheimer and fixed nuclei approximations are only made in non-quantum simulations. IRWolfie- (talk) 22:37, 25 November 2011 (UTC)

I've clarified it myself. IRWolfie- (talk) 15:07, 27 November 2011 (UTC)

Discuss reverts

P99am, instead of reverting every edit I've made do you want to justify every revert you made. Here is a diff [1] IRWolfie- (talk) 15:30, 8 December 2011 (UTC)

Sorry, but your editing does not improve the article. It looks like change for the sake of change. You add and remove something without clear motivation. For example you remove "A popular software for ab-initio molecular dynamics is the Car-Parrinello Molecular Dynamics (CPMD) package based on the density functional theory." It is not very well written, but it's valuable information and it should be here. This is the most used method. In general, there was a deterioration, not improvement of the article. (Please note that I did not revert "every edit You've made". The only useful editing is still here.) P99am (talk) 10:51, 9 December 2011 (UTC)

Mention of the Car-Parrinello method was unreferenced. The list of packages is a duplication of an existing article. Provide actual reasoning when reverting my edits rather than that you don't like it. IRWolfie- (talk) 11:02, 9 December 2011 (UTC)

The situation is reversed, you should give the reasons for your editing, which is not writing the article, but is removing the existing (and useful) information. If you insist on your version, then please refer to the community in order to prevent the Edit war. Sorry I restore the original version. P99am (talk) 11:41, 9 December 2011 (UTC)

Again; the Mention of the Car-Parrinello method is unreferenced. The list of packages is a duplication of an existing article. Provide actual reasoning when reverting my edits rather than that you don't like it. IRWolfie- (talk) 11:02, 9 December 2011 (UTC)

If you look at the two sections above this you will see comments I have made, Would you care to respond? IRWolfie- (talk) 20:43, 13 December 2011 (UTC)

The list of MD software is duplication of an existing article and unnecessary, I have hence removed it. The See also section was (and probably still) contained far too many links so I trimmed that too. See WP:SEEALSO. Polyamorph (talk) 07:48, 14 December 2011 (UTC)

This version is acceptable except for "A quantum description is still used for the dynamics of the electrons when these approximations are made." I erased it. P99am (talk) 10:33, 14 December 2011 (UTC)

Do you care to provide some reasoning? IRWolfie- (talk) 15:19, 14 December 2011 (UTC)

The absence of sense. P99am (talk) 15:40, 14 December 2011 (UTC)

Yet again you avoid giving reasoning for anything. IRWolfie- (talk) 09:35, 15 December 2011 (UTC)

If you ask my 3rd opinion, the entire paragraph about the "realm" should be removed as unsourced and not telling anything specific about the subject. Biophys (talk) 16:03, 3 March 2012 (UTC)

NPOV

I think this article has a serious issue with WP:NPOV. It does not tell much about limitations of the method. As a reader of the article, I would like to know if MD can be used for solving any practical problems, such as ligand docking, homology modeling or protein structure prediction. If it can be used, what was performance of MD based on community-wide assessments, such as CASP? The method takes a number of concepts from classical physics of macroscopic objects (even such as dielectric constant) and uses them for molecules and atoms. Does it cause any serious problems? What problems exactly? And so on. Perhaps I could add something about this if there are no objections.My very best wishes (talk) 15:45, 4 March 2012 (UTC)

Let's be more clear. The failure of MM/MD in CASP experiments (as noted by Michael Levitt, one of the first researchers who used MD for proteins) occur for at least three different reasons:

1. In any practical applications (protein folding, ligand docking) the MD is used essentially as a method of global energy optimization. Which energy? This is potential energy of a system as understood in mechanics, no matter classical or quantum. To be more precise, the "force fields" in Chemistry have been typically developed in a such way that absolute value of the potential energy reproduces the enthalpy of sublimation of molecular crystals. But the mechanical potential energy is actually irrelevant here. The processes we are talking about (protein folding, ligand docking) are governed by changes in the Gibbs free energy that also includes conformational entropy and other primarily entropic contributions, most notably the hydrophobic effect. Just to compare, the "conformational energy" (enthalpy in vacuum) difference between an extended and folded conformations of a protein is typically around several thousand kcal/mol, whereas the experimental Gibbs free energy differences for a protein are only several kcal/mol. Of course, the required free energy difference between the unfolded and folded protein conformations can be calculated from a statistical ensemble of conformers generated by MM/MD, but this is not the energy optimized during the MD simulations.
2. All interatomic/intermolecular interactions are ultimately of electrostatic origin (even dispersion attraction). Therefore, they depend on dielectric properties of the environment. This concerns not only Coulomb electrostatics and H-bonds, but also van der Waals interactions that become effectively weaker and start following "like dissolves like" rule in condensed media (unlike that in vacuum and MM/MD). This has been described, for example, in "intermolecular and surface forces" by Israelaschvilli (if we need a secondary RS). This problem was supposed to be fixed by developing "polarizable force fields", but it hardly was.
3. Calculating huge energies (thousands kcal/mol) leads to statistical accumulations of errors, making a precise calculation of very small differences (several kcal/mol) practically impossible (I saw a couple publications about this).

My very best wishes (talk) 15:45, 4 March 2012 (UTC)

I guess no one cares, so I quickly NPOVed some of that, however this needs more work and better sourcing. My very best wishes (talk) 02:54, 8 March 2013 (UTC)

I'm watching the article, I'll see what can be done when I have some time. • Jesse V.^(talk) 03:30, 8 March 2013 (UTC)

Examples of applications

This section seems like trivia for now (perhaps it could be useful if it gave a comprehensive listing of types of simulations that people run, but until then I think that the page would be better without it).

Additionally, because computers continue to get faster, listings of "very long simulations" get stale rather quickly. The simulations cited in the section following the text "The following two biophysical examples are not run-of-the-mill MD simulations" are now no longer the longest simulations or the biggest systems simulated to date. In fact, simulation studies totalling 0.5 ms are no longer particularly rare (note that this "0.5 ms simulation" noted in the article is not a single continuous simulation but the aggregated simulation time of many shorter simulations).

Finally, there is a danger that a section like this is simply an advertisement for a biased selection of work. — Preceding unsigned comment added by 108.161.122.218 (talk) 19:11, 23 September 2012 (UTC)

Merge with Molecular dynamics

The whole article is highly redundant--92.205.19.67 (talk) 10:35, 12 February 2014 (UTC) — Preceding unsigned comment added by 92.192.57.120 (talk)

• It is an opportunity to improve the article, but do not to merge. These things are considerably different. — Preceding unsigned comment added by P99am (talk • contribs)

Yep, Second Quantization (talk) 09:28, 20 February 2014 (UTC)

The flow diagram below infobox is missing velocity update

It is crucially missing v=v+a*dt. Am I right? Oceangai (talk) 23:45, 27 February 2014 (UTC)

I think you are correct. Sizeofint (talk) 03:33, 21 May 2015 (UTC)

Yes, of course it was missing it, it was a simplified one (made by me). But instead of just removing it, please make a better one then and add it there. Just removing other people stuff without providing an improved version is not very constructive. — Preceding unsigned comment added by 128.214.7.97 (talk) 06:40, 11 June 2015 (UTC)

I now added a much more detailed MD algorithm schematic -- of course this is not exactly correct either, it is impossible to make an exactly correct one in a schematic. If anybody is not happy with this, don:'t remove it, but replace it with a better one. Knordlun (talk) 09:32, 7 July 2015 (UTC)

Looks great! Thanks! Sizeofint (talk) 15:39, 7 July 2015 (UTC)

Mistaken summation?

Unless I'm misunderstanding something... in the section 'Pair potentials versus many-body potentials', in the second equation (for U sub i,j) of that section, the first term should not have a summation, right?

Sincerely, DrTLesterThomas (talk) 18:43, 19 February 2015 (UTC)

Should any of the terms have summations? If this is the potential between a pair of particles why sum over all particles? Sizeofint (talk) 03:50, 21 May 2015 (UTC)

Paragraph to remove

I propose to delete the paragraph beginning with the words "Some results of simulations ..." This section contains some amateur discussion of protein folding problem. In spite of the serious references the text is misleading. Article does not lose anything because this section is not about MD itself but about the specific problem. P99am (talk) 09:35, 21 June 2015 (UTC)

Agreed Sizeofint (talk) 15:43, 21 June 2015 (UTC)

• This has been explained in a section above. If you want to improve the description of this method limitations, that's fine - please bring more sources about this and rewrite. However, noticing limitations of the method is important per WP:NPOV. This part should not be simply removed. My very best wishes (talk) 18:34, 31 August 2015 (UTC)

  I agree that there should be discussion of the method's limitations, but the existing text is really dated. The most recent reference in that paragraph is Levitt's 1999 paper! The preceding list of biophysics applications reads strangely, too; certainly docking shouldn't be first on the list. Opabinia regalis (talk) 22:17, 2 September 2015 (UTC)

This is a complicated issue to discuss. Hence just a few points.

1. I agree that text should be probably rearranged to clarify that it was about MD in general, rather than about protein folding (protein folding is simply an example that clarifies what the general problems are);
2. Nothing in the underlying force fields which are currently used has fundamentally changed during this time to my knowledge, so all arguments and sources are still valid.
3. If you or anyone else want to bring newer publications on the subject of MD applicability (and I know that a lot has been published), they are more than welcome. For example, someone might think that polarizable force fields helped to resolve any problems described in the "limitations" section. If so, one should simply describe: (a) what the problems are/were, and (b) how exactly they have been resolved - per sources. I can not do it because to my knowledge these problems have not been resolved. If they were resolved, we would had precise 3D models of all proteins from numerous genomes generated by energy minimization and MD. But we only have very poor quality models generated at the Protein Model Portal by very old and primitive homology modeling because it works much better than "ab initio" MD. My very best wishes (talk) 01:54, 4 September 2015 (UTC)
4. P.S. If Michael Levitt changed his opinion about MD, this should be noted. My very best wishes (talk) 02:14, 4 September 2015 (UTC)

While there haven't been any fundamental changes in the basic model, there have been significant advances in the quality of the force field parameters - as judged by e.g. comparison to NMR measurements - and in the compute resources available, as well as a correspondingly deeper appreciation of protein conformational variability on the experimental side of things. None of that stuff is a game-changer on its own but collectively the relative weights of different problems have changed quite a bit with respect to application to proteins.

I don't know if Levitt himself has changed his opinion about MD, but the quote highlighted in the article really does sound like old business. Overall interest in CASP has declined, Rosetta is everywhere, and model refinement isn't really a common use case, current or anticipated, for MD. The model portal and similar 'model the proteome' style projects have always been mostly grant bait, and polarizability is vaporware.

I started to rewrite some of this but it's too much like doing work after work ;) At least the applications sentence should really be refactored; putting docking first has to be some kind of clever troll. Docking people outside of Schrödinger can't seem to stand MD. Opabinia regalis (talk) 06:44, 4 September 2015 (UTC)

That's fine. But we need a "criticism" section written in a language people can understand. OK, let's say it in a different way. People who use this method tend to think that they are seeing actual/real movements of atoms. Yes, in a way they do - in a very crude approximation, and on a limited time scale. However, method fail during simulations of complex processes, such as protein folding. That has been openly admitted by a scientist who received Nobel prize for developing and applying this method to proteins. This is happening because the underlying "potential energy" functions do not directly account for a number of fundamentally important factors, such as the environment-dependence of interatomic interactions, hydrophobic interactions and conformational entropy. That's why "docking people" usually do not use MD (actually, some of them do) - exactly as you tell. That's why people are developing alternative energy functions, even such as statistical potential. My very best wishes (talk) 13:16, 4 September 2015 (UTC)

I suppose I'm circuitously getting back around to your point 1, that a discussion of limitations should distinguish between the general limitations that arise due to the approximations made in modeling the underlying physics, and the issues that apply to specific use cases. Poor-quality force field parameters is an issue with a foot on either side of the fence, and if you believe any of the DE Shaw papers from a couple of years ago (e.g. PMID 22513870), is the most significant obstacle to both protein folding and model refinement (at least, if you have vast amounts of computing resources to throw at the issue). Opabinia regalis (talk) 04:31, 5 September 2015 (UTC)

Yes, I think this paper by Shaw is worth mentioning, even though he did this in "post-diction" (not prediction/CASP) regime. The CASPs are indeed important as the only believable way to assess the computational methods, most of which are not MD. The problems/limitations are very general but they become more apparent when one deals with global energy minimization of a complex system. In this case one should be looking for the global minimum of Gibbs free energy difference (relative to the coil in aqueous solution for proteins). Some people who do MD simulations are claiming to achieve just that, ironically with coarse-grained models. Can they do something real and better than using other methods? Only CASP can answer, but for someone who understands the physics the answer is "no". My very best wishes (talk) 12:51, 5 September 2015 (UTC)

Yes, I would suggest this review: PMID 24463371 from the Shaw group also, and a benchmark of modern force fields against NMR parameters, e.g. PMID 22754404 from Pande.

From the view of a physicist, yes, the deficiencies are general and irresolvable without modifying the model. From the view of a protein biochemist, the deficiencies are relevant only to the extent that they affect the results, which for most uses are not ab initio predictions of folded globular structures. (Arguably most usage of MD by experimentalists is as a confirmation bias engine, but that's OR... ;) Opabinia regalis (talk) 05:46, 7 September 2015 (UTC)

OK. But a statement from second paper is misleading ("Although early force field development was limited by the lack of direct comparisons between simulation and experiment ..."). This is not true. The parameters of all force fields have been actually fitted to reproduce experimental data from the Day 1. The only question which exactly data. For example, one can reproduce equilibrium geometries of molecules in the crystal if interatomic distances in 6-12 potentials are right, regardless to the depths of potentials. No question that a lot of various data can be reproduced, however if method fails in practical problems, such as docking, (and the specialized "docking CASPs" did not show good performance of MD), there should be something seriously problematic with MD. 14:30, 7 September 2015 (UTC)

• I do not have time to contribute, but just for a summary (very simplified):

1. Please note that Shaw and Pande use exactly or essentially the same force fields as 20 years ago. Nothing had really changed. Same on recent CASPs. Best performance was shown by people who used fold recognition, which is essentially a version of homology modeling. Some progress had been definitely achieved, but by developing fold recogintion methods.
2. Whatever Shaw had demonstrated, this is not really significant because the problem is not the time of simulation, but the underlying "energy", exactly as Michael Levitt and some others (including Shaw) had noted in their publications. The key finding is that MD was not really useful in the folding and docking CASPs.
3. This begs the question: why? And the answer is pretty obvious: a number of key physical factors were not taken into account in MM/MD (which of course has been discussed in the literature). This is very general.
4. It also helps to know the "kitchen": how exactly the parameters of the energy functions have been derived. In fact, they have been derived by fitting. This is no "better" than parameters in implicit solvent models or QSAR. The only difference is which parameters have been fitted. If one is interested in the enthalpy of sublimation, then yes, one should use MM/MD force fields (thousands of kcal/mol for a protein in CHARMM etc.). However, the enthalpy of sublimation is irrelevant for protein folding, ligand docking and practically any processes that occur in the aqueous solution, or more generally in condensed media. The relevant are Gibbs free energies differences (just a few kcal/mol for protein folding or binding). If one is interested in a fast calculation/prediction of relevant energy, then using the implicit solvent models which are based on direct fitting of transfer free energies, or QSAR models that are based on direct fitting of ligand binding free energies is the way to go. That's why many "docking people" are using versions of QSAR. It simply works better, just as fold recognition works better - as expected. My very best wishes (talk) 12:16, 9 September 2015 (UTC)

Oops, I fell behind on this, sorry. I made a few tweaks, but don't have time for a large-scale change either. To respond to the above, the "experimental data" almost universally intended by the relevant literature is data on full-length proteins, usually NMR observables, not related to the data used to fit the force field parameters. I still think there is too much emphasis on CASP performance, which is not what most users of MD care very much about, and whose perceived relevance has faded substantially. Funny, in my mind QSAR is one of those things that never really works as well as it's supposed to ;) Opabinia regalis (talk) 06:16, 10 September 2015 (UTC)

OK, I have no objections to your changes. Maybe I will try to improve this part slightly in a future. I too would not argue in favor of QSAR, but it is widely used in fields like medicinal chemistry, and it is actually working, although with a number of serious restrictions that belong to page QSAR... My very best wishes (talk) 22:25, 10 September 2015 (UTC)

A single run of an MD simulation optimizes the potential energy, rather than the free energy of the protein, meaning that all entropic contributions to thermodynamic stability of protein structure are neglected.

The thermodynamics effects are usually considered because the thermal motion of the protein is actually simulated. — Preceding unsigned comment added by 146.193.56.115 (talk) 08:30, 21 January 2016 (UTC)

Do you want this included in the article somewhere? If so we need a reliable source. Sizeofint (talk) 17:02, 21 January 2016 (UTC)