Mohammad M. Sultan; Vijay S. Pande
2017-05-09 (online) – 2017-06-13 (print)
Journal of Chemical Theory and Computation (J. Chem. Theory Comput.). 13, 6, 2440-2447. doi:10.1021/acs.jctc.7b00182
Hao Wu; Feliks Nüske; Fabian Paul; Stefan Klus; Péter Koltai; Frank Noé
2017-04-21 (print)
The Journal of Chemical Physics (J. Chem. Phys.). 146, 15, 154104. doi:10.1063/1.4979344 arxiv:1610.06773
Provides a better way of "symetrizing" tICA correlation matrix. In tICA, you assume that the dynamics are reversible. When we're learning from finite data, this reversibility isn't respected. Historically, you take your correlation matrix, add its transpose, and divide by two. This is an especially poor approximation if you have many short trajectories. This paper is analogous to the MLE method for symetrizing MSM counts matrices.
Show References| Num | Entry | Why |
|---|---|---|
| 46 | 2015-uncertainty-estimation | Reversibility makes analysis easier |
| 50 | 2013-noe-variational | |
| 51 | 2013-noe-tica | |
| 52 | 2014-nuske-variational |
msm-theory
Marcus Weber; Konstantin Fackeldey; Christof Schütte
2017-03-28 (print)
The Journal of Chemical Physics (J. Chem. Phys.). 146, 12, 124133. doi:10.1063/1.4978501
Collection of m "base points" they make a gaussian RBF of distances to base points. They normalize it to unity. This is the softmax function, but they don't call it that.
They add base points adaptively and use PCCA+ to lump.
Note that they have stopped calling this "meshless" or "mesh-free", probably because the regular MSM is also meshless. Now the abstract says "This kind of meshless discretization..."
Robert T. McGibbon; Brooke E. Husic; Vijay S. Pande
2017-01-28 (print)
The Journal of Chemical Physics (J. Chem. Phys.). 146, 4, 044109. doi:10.1063/1.4974306 arxiv:1602.08776
The authors argue for a definition of the reaction coordinate as the projection on the dominant eigenfunciton of the propogator. Notably, they say that path-based coordinates are no good, because progress is only defined along the path. They argue that the coordinate shouldn't depend on start and end points. They say the projection should be maximally predictive. This means finding the slowest modes. They note 2006-nadler-diffusion-maps (ref. 61) and 2011-rohrdanz-diffusion-maps (ref. 62) have used this definition.
They go on to show tICA finds this reaction coordinate. To make tICA more interpretable, they develop an algorithm for introducing a sparsity pattern. It's a pseudo-l0 regularization (made smooth so the optimization works).
They also use a unique dihedral featurization: instead of taking the sine and cosine to get around periodicity concerns; they project the values on a bunch of evenly spaced von-mises (periodic gaussians) distributions around the unit circle. Each dihedral is expanded into several numbers. It's like a smooth histogramming. This probably won't work as the number of dihedrals gets large (too many features).
Show References| Num | Entry | Why |
|---|---|---|
| 61 | 2006-nadler-diffusion-maps | |
| 62 | 2011-rohrdanz-diffusion-maps |
msm-theory tica features
Brooke E. Husic; Robert T. McGibbon; Mohammad M. Sultan; Vijay S. Pande
2016-11-21 (print)
The Journal of Chemical Physics (J. Chem. Phys.). 145, 19, 194103. doi:10.1063/1.4967809
The authors perform GMRQ cross validation on the twelve 2011-larsen-folding folding trajectories to give guidelines for MSM construction.
They present a flowchart for MSM construction that shows the three paths towards clustering: from an rmsd distance metric, from features, or from tICA learned on features.
They introduce GMRQ cross validation in the tradition of 2015-mcgibbon-gmrq (ref. 44).
They present results but stress that you have to do your own cross validataion to be sure. Some conclusions include: 1. tICA and PCA are better than direct clustering of features 2. when using tica, you can use kcenters, kmeans, or minibatch kmeans to the same effect
On one protein (2p6j) they look at all different features and show that they vary a lot. It's unfortunate that this was only done on one protein.
Show References| Num | Entry | Why |
|---|---|---|
| 41 | 2013-noe-variational | Variational principle |
| 42 | 2014-nuske-variational | Variational principle |
| 5 | 2008-anton | Generated the trajectories. |
| 18 | 2011-larsen-folding | Re-analyzed these simulations. "Diversity of proteins analyzed" |
| 44 | 2015-mcgibbon-gmrq | Cross-validation |
msm-theory
Frank Noé; Ralf Banisch; Cecilia Clementi
2016-11-08 (print)
Journal of Chemical Theory and Computation (J. Chem. Theory Comput.). 12, 11, 5620-5630. doi:10.1021/acs.jctc.6b00762
Scale tIC coordinates by a function of the timescale. See also 2015-kinetic-mapping.
msm-theory
Robert McGibbon
2016-02-12 (online)
Benjamin Trendelkamp-Schroer; Hao Wu; Fabian Paul; Frank Noé
2015-11-07 (print)
The Journal of Chemical Physics (J. Chem. Phys.). 143, 17, 174101. doi:10.1063/1.4934536
Reversible estimates for MSMs
F. Vitalini; F. Noé; B. G. Keller
2015-09-08 (print)
Journal of Chemical Theory and Computation (J. Chem. Theory Comput.). 11, 9, 3992-4004. doi:10.1021/acs.jctc.5b00498
Authors simulate individual (capped) amino acids for 1us / each and construct (mini-)MSMs on each one. They use the outerproduct of these mini-MSMs to serve as a basis set for peptides. MiniMSMs are on a grid in phi-psi angles. Since each miniMSM has approx 3 modes, the full basis would be 3^(N), which is way too big! They call the second and third modes "excited states" and use a basis set that contains a singly-exited residue. E.g. 11111 + [ [21111, 121111, 112111, 111211, ...] ].
Alanine preceded by a proline is taken as a special case.
msm-theory
Robert T. McGibbon; Vijay S. Pande
2015-07-21 (print)
The Journal of Chemical Physics (J. Chem. Phys.). 143, 3, 034109. doi:10.1063/1.4926516
msm-theory
Ch. Schütte; M. Sarich
2015-06-22 (online) – 2015-09-01 (print)
The European Physical Journal Special Topics (The European Physical Journal Special Topics). 224, 12, 2445-2462. doi:10.1140/epjst/e2015-02421-0
Transfer operator 1999-schutte-msm (ref. 1).
Coarse grain MSM states 2000-pcca (ref. 2) 2005-pcca (ref. 3).
Meshless MSMs 2006-meshless-msm-thesis (ref. 24) 2011-meshless-msm (ref. 32) 2011-meshless-msm (ref. 33). Wikipedia says these are also called "meshfree" methods.
Show References| Num | Entry | Why |
|---|---|---|
| 1 | 1999-schutte-msm | |
| 2 | 2000-pcca | |
| 3 | 2005-pcca | |
| 24 | 2006-meshless-msm-thesis | |
| 32 | 2011-meshless-msm | |
| 33 | 2011-meshless-msm |
Frank Noe; Cecilia Clementi
2015-06-20 (online)
Scale tIC coorindates by the eigenvalue. See also 2016-commute-maps.
Robert T. McGibbon; Vijay S. Pande
2015-03-28 (print)
The Journal of Chemical Physics (J. Chem. Phys.). 142, 12, 124105. doi:10.1063/1.4916292
msm-theory variational
Matthew P. Harrigan; Diwakar Shukla; Vijay S. Pande
2015-03-10 (print)
Journal of Chemical Theory and Computation (J. Chem. Theory Comput.). 11, 3, 1094-1101. doi:10.1021/ct5010017
Solvent-shells featurization for including solvent in MSM construction.
Hao Wu; Frank Noé
2015-02-28 (print)
The Journal of Chemical Physics (J. Chem. Phys.). 142, 8, 084104. doi:10.1063/1.4913214
Variational method 2013-noe-variational (ref. 26) 2014-nuske-variational (ref. 27).
MSM is variational with step functions 2013-noe-variational (ref. 26).
"Markov transition models (MTMs)", specifically Gaussian mixtures (GMTM).
Show References| Num | Entry | Why |
|---|---|---|
| 26 | 2013-noe-variational | |
| 27 | 2014-nuske-variational | |
| 26 | 2013-noe-variational |
msm-theory
Diwakar Shukla; Carlos X. Hernández; Jeffrey K. Weber; Vijay S. Pande
2015-02-17 (print)
Accounts of Chemical Research (Acc. Chem. Res.). 48, 2, 414-422. doi:10.1021/ar5002999
review
Christian R. Schwantes; Vijay S. Pande
2015-02-10 (print)
Journal of Chemical Theory and Computation (J. Chem. Theory Comput.). 11, 2, 600-608. doi:10.1021/ct5007357
They introduce kernel tICA as an extension to tICA. This is useful to get non-linear solutions to the tICA equation. They claim you can estimate eigenprocesses without building an MSM.
They briefly introduce the transfer operator. They introduce the variational principle of conformation dynamics per 2011-prinz (ref. 25). They introduce tICA as maximizing the autocorrelation. They say that solutions to tICA are the same as solutions to the variational problem per 2013-noe-tica (ref. 28). Linearity makes them crude solutions.
They explain that a natural approach to introduce non-linearity is to expand the original representation into a higher dimensional space and do tICA there. They say this is impractical. The expanded space probably has to be huge. You can perform analysis in the big representation without explicitly representing it by using the "kernel trick". They reproduce an example of the kernel trick from 1998-scholkopf-kernel-pca (ref. 39).
They re-write the tICA problem only in terms of inner products so you can apply the kernel trick. They introduce normalization. They choose a gaussian kernel. They simulate a four-well potential, muller potential, alanine dipeptide, and fip35ww. They need to do MLE cross validation over parameters (kernel width and regularization strength).
This uses so much RAM! Huge matrices to solve (that scale with the amount of data!!)
Show References| Num | Entry | Why |
|---|---|---|
| 21 | 2014-msm-perspective | Data needs analysis |
| 25 | 2011-prinz | Details of transfer operator approach. |
| 33 | 2001-schutte-variational | Details of transfer operator approach. |
| 34 | 2013-noe-variational | "It was shown that a variational principle can be derived for the eignvalues of the transfer operator." The autocorelation of a function is less than the autocorrelation of the first dynamical eigenfunction of the transfer operator. This is used to argue that you don't have to estimate the operator itself. Just estimate its eigenfunctions |
| 35 | 2014-nuske-variational | "Successfully constructed estimates of the top eigenfunctions in the span of a prespecified library of basis functions." Contrast with this work, which "does not require a predefined basis set" |
| 22 | 2013-schwantes-tica | Citing tICA |
| 28 | 2013-noe-tica | solutions to tica provide estimates of the slowest eigenfunctions of the transfer operator. |
| 36 | doi:10.1103/PhysRevLett.72.3634 | Citing tICA |
| 37 | doi:10.1162/neco.2006.18.10.2495 | Citing tICA |
| 39 | 1998-scholkopf-kernel-pca | Used to introduce ther kernel trick. |
msm-theory
C. R. Schwantes; R. T. McGibbon; V. S. Pande
2014-09-07 (print)
The Journal of Chemical Physics (J. Chem. Phys.). 141, 9, 090901. doi:10.1063/1.4895044
Very good perspective on the importance of analysis (particularly MSM analysis) for understanding large, modern MD datasets. Money quote: "we believe that quantitative analysis has increasingly become a limiting factor in the application of MD"
msm-theory perspective
Robert T. McGibbon; Christian R. Schwantes; Vijay S. Pande
2014-06-19 (print)
The Journal of Physical Chemistry B (J. Phys. Chem. B). 118, 24, 6475-6481. doi:10.1021/jp411822r
This is before 2015-mcgibbon-gmrq GRMQ cross-validation. They explicitly find the volume of voronoi cells (in low number of tIC space) to find a likelihood. They use AIC/BIC to find the number of states to use. Finding volumes is tough and you still can't compare across protocols (so you can basically only scan number of states or clustering method), but! this was the first paper to seriously suggest using a smaller number of states to avoid overfitting.
msm cross-validation
Feliks Nüske; Bettina G. Keller; Guillermo Pérez-Hernández; Antonia S. J. S. Mey; Frank Noé
2014-04-08 (print)
Journal of Chemical Theory and Computation (J. Chem. Theory Comput.). 10, 4, 1739-1752. doi:10.1021/ct4009156
This paper is largely redundant with 2013-noe-variational (ref. 65). They cite it as such: "Following the recently introduced variational principle for metastable stochastic processes,(65) we propose a variational approach to molecular kinetics."
They perform their variational approach on 2- and 10-alanine in addition to 1D potentials.
This comes after tICA and cites 2013-schwantes-tica (ref. 57) and 2013-noe-tica (ref. 58) in the intro, but does nothing further with it. In particular, they don't note that tICA is just another choice of basis set.
They cite their error paper 2010-msm-error (ref. 55).
Show References| Num | Entry | Why |
|---|---|---|
| 65 | 2013-noe-variational | |
| 57 | 2013-schwantes-tica | |
| 58 | 2013-noe-tica | |
| 55 | 2010-msm-error |
msm-theory variational
John D Chodera; Frank Noé
2014-04-01 (print)
Current Opinion in Structural Biology (Curr. Opin. Struct. Biol.). 25, 135-144. doi:10.1016/j.sbi.2014.04.002
Overview of MSMs, stressing eigensystem and variational approach. Includes further reading suggestions.
msm-theory perspective
Jan-Hendrik Prinz; John D. Chodera; Frank Noé
2014-02-21 (online)
Physical Review X (Physical Review X). 4, 1, doi:10.1103/PhysRevX.4.011020
Kai J. Kohlhoff; Diwakar Shukla; Morgan Lawrenz; Gregory R. Bowman; David E. Konerding; Dan Belov; Russ B. Altman; Vijay S. Pande
2013-12-15 (online)
Nature Chemistry (Nature Chem.). 6, 1, 15-21. doi:10.1038/nchem.1821
They used Google's Exacycle to do these simulations. You can cite this for more examples of distributed computing. It's ostensibly about GPCRs.
distributed-computing
Frank Noé; Hao Wu; Jan-Hendrik Prinz; Nuria Plattner
2013-11-14 (print)
The Journal of Chemical Physics (J. Chem. Phys.). 139, 18, 184114. doi:10.1063/1.4828816
Wenwei Zheng; Mary A. Rohrdanz; Cecilia Clementi
2013-10-24 (print)
The Journal of Physical Chemistry B (J. Phys. Chem. B). 117, 42, 12769-12776. doi:10.1021/jp401911h
Use diffusion maps to run umberlla sampling
Robert T. McGibbon; Vijay S. Pande
2013-07-09 (print)
Journal of Chemical Theory and Computation (J. Chem. Theory Comput.). 9, 7, 2900-2906. doi:10.1021/ct400132h
Learn scaling of coordinates to better approximate kinetics? Redundant with tICA.
Guillermo Pérez-Hernández; Fabian Paul; Toni Giorgino; Gianni De Fabritiis; Frank Noé
2013-07-07 (print)
The Journal of Chemical Physics (J. Chem. Phys.). 139, 1, 015102. doi:10.1063/1.4811489
The Noe group introduces tica concomitantly with 2013-schwantes-tica. They use the variational approach from 2013-noe-variational to derive the tICA equation. They cite a 2001 book about independent component analysis.
msm-theory tica
Christian R. Schwantes; Vijay S. Pande
2013-04-09 (print)
Journal of Chemical Theory and Computation (J. Chem. Theory Comput.). 9, 4, 2000-2009. doi:10.1021/ct300878a
The Pande group introduces tica concomitantly with 2013-noe-tica. This paper uses PCA as inspiration and cites signal processing literature.
msm-theory tica
Thomas J Lane; Diwakar Shukla; Kyle A Beauchamp; Vijay S Pande
2013-02-01 (print)
Current Opinion in Structural Biology (Curr. Opin. Struct. Biol.). 23, 1, 58-65. doi:10.1016/j.sbi.2012.11.002
The state of folding simulations as it was in 2013. Has a nice plot of folding time by year by lab. Discusses the state of MSMs for analysis. Maybe cite this if you're doing folding or want to talk about how timescales are getting longer (and analysis is getting harder). The references include "recommended readings", which is nice.
msm-theory perspective
Frank Noé; Feliks Nüske
2013-01-01 (print)
Multiscale Modeling & Simulation (Multiscale Model. Simul.). 11, 2, 635-655. doi:10.1137/110858616
I think the point of this versus 2014-nuske-variational is to be "protein agnostic". They allude to proteins, but say this is more general. Their example is a double-well potential.
They introduce the propogator formalism and stipulate that dynamics can be seperated into "fast" and "slow" components. In contrast to a quantum mechanics Hamiltonian, we don't know the propogator here. You have to infer it from data.
They claim the error bound derived in 2010-msm-error (ref. 34) is not constructive, whereas this method *is* constructive.
Math section heavily cites 2010-msm-error (ref. 34).
They adapt the Rayleigh variational principle from quantum mechanics, and cite 1989-szabo-ostlund-qm (ref. 43). They show that the autocorrelation of the true first dynamical eigenfunction is its eigenvalue, and an estimate of the first dynamical eigenfunction necessarily has a smaller eigenvalue. This sets the variational bound. In terms of names that don't seem to be used now that we're in the future: the Ritz method is for when you have no overlap integrals (e.g. MSMs) and the Roothan-Hall method is for when you do (tICA).
They put it to the test on a double well potential. They use indicator basis functions to make an MSM; hermite basis functions so they still have no overlap integrals, but smooth functions; and gaussian basis functions (with overlap integrals). This must have come before tICA because there is no mention made of it, even though it would fit in nicely.
Show References| Num | Entry | Why |
|---|---|---|
| 34 | 2010-msm-error | |
| 34 | 2010-msm-error | |
| 43 | 1989-szabo-ostlund-qm |
msm-theory variational
Konstantin Fackeldey; Alexander Bujotzek; Marcus Weber
2012-09-27 (online) – 2013-01-01 (print)
Lecture Notes in Computational Science and Engineering (Lect. Notes Comput. Sci. Eng.). 141-154. doi:10.1007/978-3-642-32979-1_9
Soften the hard clustering 2006-meshless-msm-thesis (ref. 37).
Cite Shepard's approach 1968-shepard-method (ref. 30) like 2006-meshless-msm-thesis does to introduce the softmax function as basis functions with softness parameter alpha. Note that this is not Shepard's method.
They frame everything in the context of lumping and PCCA+ and use ZIBgridfree to simulate trialanine faster than unbiased (100ns vs 10ns).
Show References| Num | Entry | Why |
|---|---|---|
| 37 | 2006-meshless-msm-thesis | |
| 30 | 1968-shepard-method |
Ting Zhou; Amedeo Caflisch
2012-08-14 (print)
Journal of Chemical Theory and Computation (J. Chem. Theory Comput.). 8, 8, 2930-2937. doi:10.1021/ct3003145
A unique featurization that encodes each atom by the first ~3 moments of its distribution of 1/distance to every other atom. Cite this if you use this featurization.
features
Tianbao Yang; Yu-Feng Li; Mehrdad Mahdavi; Rong Jin; Zhi-Hua Zhou
2012-01-01 (print)
Advances in Neural Information Processing Systems (Advances in Neural Information Processing Systems). 476-484.
Natasa Djurdjevac; Marco Sarich; Christof Schütte
2012-01-01 (print)
Multiscale Modeling & Simulation (Multiscale Model. Simul.). 10, 1, 61-81. doi:10.1137/100798910
Konstantin Fackeldey; Susanna Röblitz; Olga Scharkoi; Marcus Weber
2011-06-22 (online)
They note MSM is a meshfree method with characteristic basis functions.
They define a "hard decomposition" in the obvious way. They define a "soft decomposition" also as a partitioning of unity, but allowing overlap.
They still do PCCA+ and it's unclear what shape function they're using for softness. As an example, they lump 504 soft states into 5 macrostates of alanine dipeptide, sometimes spelled alanin dipeptid
Jan-Hendrik Prinz; Hao Wu; Marco Sarich; Bettina Keller; Martin Senne; Martin Held; John D. Chodera; Christof Schütte; Frank Noé
2011-05-07 (print)
The Journal of Chemical Physics (J. Chem. Phys.). 134, 17, 174105. doi:10.1063/1.3565032
Fantastic in-depth intro to MSMs. Figure 1 in this paper is necessary for understanding eigenvectors. This defines and relates the propogator and transfer operator. This shows how we compute timescales from eigenvectors. This discusess state decomposition error and shows that many states are needed in transition regions.
quote: it is clear that a “sufficiently fine” partitioning will be able to resolve “sufficient” detail 2010-msm-error.
Cites 2004-nina-msm for use of the term "MSM".
msm-theory review
Mary A. Rohrdanz; Wenwei Zheng; Mauro Maggioni; Cecilia Clementi
2011-03-28 (print)
The Journal of Chemical Physics (J. Chem. Phys.). 134, 12, 124116. doi:10.1063/1.3569857
other-md-analysis
Yusuke Naritomi; Sotaro Fuchigami
2011-02-14 (print)
The Journal of Chemical Physics (J. Chem. Phys.). 134, 6, 065101. doi:10.1063/1.3554380
Probably the first application of tICA to MD.
msm-theory tica
Vijay S. Pande
2010-11-05 (online)
Physical Review Letters (Phys. Rev. Lett.). 105, 19, doi:10.1103/PhysRevLett.105.198101
Non-native interactions and misfolding
Peter L. Freddolino; Christopher B. Harrison; Yanxin Liu; Klaus Schulten
2010-10-01 (online) – 2010-10-01 (print)
Nature Physics (Nat. Phys.). 6, 10, 751-758. doi:10.1038/nphys1713
Cited by 2014-msm-perspective as highlighting analysis as a problem.
Vijay S. Pande; Kyle Beauchamp; Gregory R. Bowman
2010-09-01 (print)
Methods (Methods). 52, 1, 99-105. doi:10.1016/j.ymeth.2010.06.002
Review of MSMs intended for "non-experts". Obviously a little dated by now.
msm-theory review
I. Buch; M. J. Harvey; T. Giorgino; D. P. Anderson; G. De Fabritiis
2010-03-22 (print)
Journal of Chemical Information and Modeling (J. Chem. Inf. Model.). 50, 3, 397-403. doi:10.1021/ci900455r
GPUGRID intro paper. Cite this alongside FAH. They (probably) did GPU distributed computing before FAH.
distributed-computing
Marco Sarich; Frank Noé; Christof Schütte
2010-01-01 (print)
Multiscale Modeling & Simulation (Multiscale Model. Simul.). 8, 4, 1154-1177. doi:10.1137/090764049
msm-theory
2009-12-01 (print)
Nature Nanotechnology (Nat. Nanotechnol.). 4, 12, 781-781. doi:10.1038/nnano.2009.356
Editorial about 1960-plenty-of-room-at-the-bottom.
Mark S. Friedrichs; Peter Eastman; Vishal Vaidyanathan; Mike Houston; Scott Legrand; Adam L. Beberg; Daniel L. Ensign; Christopher M. Bruns; Vijay S. Pande
2009-04-30 (print)
Journal of Computational Chemistry (J. Comput. Chem.). 30, 6, 864-872. doi:10.1002/jcc.21209
Probably the second instance of using GPUs for molecular dynamics. This became OpenMM.
md-sampling
John L Klepeis; Kresten Lindorff-Larsen; Ron O Dror; David E Shaw
2009-04-01 (print)
Current Opinion in Structural Biology (Curr. Opin. Struct. Biol.). 19, 2, 120-127. doi:10.1016/j.sbi.2009.03.004
Peter Eastman; Vijay S. Pande
2009-01-01 (print)
Journal of Computational Chemistry (J. Comput. Chem.). NA-NA. doi:10.1002/jcc.21413
Optimizing below-cutoff nonbonded calculations on the GPU by tricky memory and parallelization management. This was for OpenMM. This is not PME.
Show References| Num | Entry | Why |
|---|
md-sampling md-algorithm
David E. Shaw; Jack C. Chao; Michael P. Eastwood; Joseph Gagliardo; J. P. Grossman; C. Richard Ho; Douglas J. Lerardi; István Kolossváry; John L. Klepeis; Timothy Layman; Christine McLeavey; Martin M. Deneroff; Mark A. Moraes; Rolf Mueller; Edward C. Priest; Yibing Shan; Jochen Spengler; Michael Theobald; Brian Towles; Stanley C. Wang; Ron O. Dror; Jeffrey S. Kuskin; Richard H. Larson; John K. Salmon; Cliff Young; Brannon Batson; Kevin J. Bowers
2008-07-01 (print)
Communications of the ACM (Commun. ACM). 51, 7, 91. doi:10.1145/1364782.1364802
The seminal Anton paper. Cite this when talking about single, long trajectories or special-purpose hardware.
md-sampling
Ken A. Dill; S. Banu Ozkan; M. Scott Shell; Thomas R. Weikl
2008-06-01 (print)
Annual Review of Biophysics (Annu. Rev. Biophys.). 37, 1, 289-316. doi:10.1146/annurev.biophys.37.092707.153558
Joshua A. Anderson; Chris D. Lorenz; A. Travesset
2008-05-01 (print)
Journal of Computational Physics (J. Comput. Phys.). 227, 10, 5342-5359. doi:10.1016/j.jcp.2008.01.047
They claim to be the first GPU accelerated MD engine too! Probably led to HOOMD, although they don't call it that in the paper.
md-sampling
John E. Stone; James C. Phillips; Peter L. Freddolino; David J. Hardy; Leonardo G. Trabuco; Klaus Schulten
2007-01-01 (online) – 2007-12-01 (print)
Journal of Computational Chemistry (J. Comput. Chem.). 28, 16, 2618-2640. doi:10.1002/jcc.20829
(Probably) the first GPU accelerated MD paper. This is for NAMD.
md-sampling
Boaz Nadler; Stéphane Lafon; Ronald R. Coifman; Ioannis G. Kevrekidis
2006-07-01 (print)
Applied and Computational Harmonic Analysis (Appl. Comput. Harmon. Anal.). 21, 1, 113-127. doi:10.1016/j.acha.2005.07.004
other-md-analysis
Marcus Weber
2006-02-01 (print)
Mainly concerned with lumping (PCCA) and setting up an iterative sampling scheme, released as ZIBgridfree.
Partition of unity using Shepard's method 1968-shepard-method (ref. 117). Definition 4.8 says these need to be positive (greater than zero) which rules out traditional MSMs. Why?
Highlights importants of softness parameter of the shape function, which they call alpha. They say Shepard's method with gaussian RBFs can be seen as a generalized Voronoi Tessellation.
Show References| Num | Entry | Why |
|---|---|---|
| 117 | 1968-shepard-method |
Peter Deuflhard; Marcus Weber
2005-03-01 (print)
Linear Algebra and its Applications (Linear Algebra Appl.). 398, 161-184. doi:10.1016/j.laa.2004.10.026
PCCA group states based on an MSM transition matrix. Specifically, it uses the eigenspectrum to do the lumping. Cite this in the methods section of your paper if you use PCCA or PCCA+.
msm-theory msm-postprocessing
William C. Swope; Jed W. Pitera; Frank Suits
2004-05-01 (print)
The Journal of Physical Chemistry B (J. Phys. Chem. B). 108, 21, 6571-6581. doi:10.1021/jp037421y
The first MSM paper. Gets pretty much everything right. Except they're convinced that you need to do state exploration via NVT or NPT and then calculate transitions by launching bespoke NVE simulations. Obviously, we just run big NPT runs and use that for both state space exploration and counting transitions.
msm-theory
Nina Singhal; Christopher D. Snow; Vijay S. Pande
2004-01-01 (print)
The Journal of Chemical Physics (J. Chem. Phys.). 121, 1, 415. doi:10.1063/1.1738647
Christopher K. I. Williams; Matthias Seeger
2001-01-01 (print)
Advances in Neural Information Processing Systems (Advances in Neural Information Processing Systems). 13, 682-688.
Ch. Schütte; W. Huisinga; P. Deuflhard
2001-01-01 (print)
Ergodic Theory, Analysis, and Efficient Simulation of Dynamical Systems (Ergodic Theory, Analysis, and Efficient Simulation of Dynamical Systems). 191-223. doi:10.1007/978-3-642-56589-2_9
Full treatment of transfer operator / propagator and build an MSM for a small RNA chain.
msm-theory
M. Shirts
2000-12-08 (print)
Science (Science). 290, 5498, 1903-1904. doi:10.1126/science.290.5498.1903
The seminal Folding at Home paper. Cite this whenever you talk about distributed computing or Folding at Home.
SETI@Home and distributed.net came before this.
distributed-computing
P. Deuflhard; W. Huisinga; A. Fischer; Ch. Schütte
2000-08-01 (print)
Linear Algebra and its Applications (Linear Algebra Appl.). 315, 1-3, 39-59. doi:10.1016/S0024-3795(00)00095-1
Ch Schütte; A Fischer; W Huisinga; P Deuflhard
1999-05-01 (print)
Journal of Computational Physics (J. Comput. Phys.). 151, 1, 146-168. doi:10.1006/jcph.1999.6231
Maybe the first time conformations were discretized and a Markov operator was made.
Bernhard Schölkopf; Alexander Smola; Klaus-Robert Müller
1998-07-01 (print)
Neural Computation (Neural Comput.). 10, 5, 1299-1319. doi:10.1162/089976698300017467
H Frauenfelder; S. Sligar; P. Wolynes
1991-12-13 (print)
Science (Science). 254, 5038, 1598-1603. doi:10.1126/science.1749933
Cited by 2011-prinz to say that there are many metastable states and many timescales.
Attila Szabo; Neil S. Ostlund
1989-01-01 (print)
Cited by 2013-noe-variational for Rayleigh variational method.
qm
Robert McGibbon; Bharath Ramsundar; Mohammad Sultan; Gert Kiss; Vijay Pande
32, 2, 1197-1205.
Use hidden markov models instead of discrete state MSMs.
N G Van-Kampen
R. Zwanzig; A. Szabo; B. Bagchi
Proceedings of the National Academy of Sciences (Proc. Natl. Acad. Sci. U.S.A.). 89, 1, 20-22.
Conformational space is huge, but proteins can fold very fast.
Matthew P Harrigan; Vijay S Pande
bioRxiv (bioRxiv).
C Levinthal
1968-01-01 (print)
(J. Chim. Phys. Physico-Chim. Biol.). 65, 44-45.
Taken from Nature's protein folding focus
Among the most widely cited-yet least read-papers in the field, partly owing to the difficulties in getting hold of them, Cyrus Levinthal used a simple model to show that a typical polypeptide chain cannot fold through an unbiased search of all conformational space on a reasonable timescale. This is commonly referred to as the "Levinthal's paradox", and led to the concept that proteins fold along discrete pathways. The first paper presents this idea and is usually cited, but the model is actually presented in the second one. Although the model was later shown to be overly simplistic, the work had a crucial role in directing the search and characterization of intermediate states.
Donald Shepard
1968-01-01 (print)
Proceedings of the 1968 23rd ACM national conference on - (Proceedings of the 1968 23rd ACM national conference on -). doi:10.1145/800186.810616
Method for interpolation confusingly cited by 2006-meshless-msm-thesis. I guess he introduces weightedsum(inverse distances) / sum(inverse distances). And instead of inverse distances, you can choose whatever function you want.
Richard Feynman
1960-02-01 (print)
Engineering and Science (Engineering and Science). 23, 5, 22-36.