Systematic reduction of the dimensionality is highly demanded in making a comprehensive interpretation of experimental and simulation data. Principal component analysis (PCA) is a widely used technique for reducing the dimensionality of molecular dynamics (MD) trajectories, which assists our understanding of MD simulation data. Here, we propose an approach that incorporates time dependence in the PCA algorithm. In the standard PCA, the eigenvectors obtained by diagonalizing the covariance matrix are time independent. In contrast, they are functions of time in our new approach, and their time evolution is implemented in the framework of Car–Parrinello or Born–Oppenheimer type adiabatic dynamics. Thanks to the time dependence, each of the step-by-step structural changes or intermittent collective fluctuations is clearly identified, which are often keys to provoking a drastic structural transformation but are easily masked in the standard PCA. The time dependence also allows for reoptimization of the principal components (PCs) according to the structural development, which can be exploited for enhanced sampling in MD simulations. The present approach is applied to phase transitions of a water model and conformational changes of a coarse-grained protein model. In the former, collective dynamics associated with the dihedral-motion in the tetrahedral network structure is found to play a key role in crystallization. In the latter, various conformations of the protein model were successfully sampled by enhancing structural fluctuation along the periodically optimized PC. Both applications clearly demonstrate the virtue of the new approach, which we refer to as time-dependent PCA.

1.
H.
Sidky
,
W.
Chen
, and
A. L.
Ferguson
, “
Machine learning for collective variable discovery and enhanced sampling in biomolecular simulation
,”
Mol. Phys.
118
,
e1737742
(
2020
).
2.
C.
Abrams
and
G.
Bussi
, “
Enhanced sampling in molecular dynamics using metadynamics, replica-exchange, and temperature-acceleration
,”
Entropy
16
,
163
199
(
2014
).
3.
H.
Fujisaki
,
K.
Moritsugu
,
Y.
Matsunaga
,
T.
Morishita
, and
L.
Maragliano
, “
Extended phase-space methods for enhanced sampling in molecular simulations: A review
,”
Front. Bioeng. Biotechnol.
3
,
125
(
2015
).
4.
C. F.
Abrams
and
E.
Vanden-Eijnden
, “
Large-scale conformational sampling of proteins using temperature-accelerated molecular dynamics
,”
Proc. Natl. Acad. Sci. U. S. A.
107
,
4961
4966
(
2010
).
5.
A. L.
Ferguson
,
A. Z.
Panagiotopoulos
,
P. G.
Debenedetti
, and
I. G.
Kevrekidis
, “
Systematic determination of order parameters for chain dynamics using diffusion maps
,”
Proc. Natl. Acad. Sci. U. S. A.
107
,
13597
13602
(
2010
).
6.
H.
Niu
,
P. M.
Piaggi
,
M.
Invernizzi
, and
M.
Parrinello
, “
Molecular dynamics simulations of liquid silica crystallization
,”
Proc. Natl. Acad. Sci. U. S. A.
115
,
5348
5352
(
2018
).
7.
P. M.
Piaggi
and
M.
Parrinello
, “
Predicting polymorphism in molecular crystals using orientational entropy
,”
Proc. Natl. Acad. Sci. U. S. A.
115
,
10251
10256
(
2018
).
8.
S. B.
Kim
,
C. J.
Dsilva
,
I. G.
Kevrekidis
, and
P. G.
Debenedetti
, “
Systematic characterization of protein folding pathways using diffusion maps: Application to Trp-cage miniprotein
,”
J. Chem. Phys.
142
,
085101
(
2015
).
9.
F.
Sicard
and
P.
Senet
, “
Reconstructing the free-energy landscape of Met-enkephalin using dihedral principal component analysis and well-tempered metadynamics
,”
J. Chem. Phys.
138
,
235101
(
2013
).
10.
T.
Ichiye
and
M.
Karplus
, “
Collective motions in proteins: A covariance analysis of atomic fluctuations in molecular dynamics and normal mode simulations
,”
Proteins: Struct., Funct., Bioinf.
11
,
205
(
1991
).
11.
A.
Kitao
and
N.
Go
, “
Investigating protein dynamics in collective coordinate space
,”
Curr. Opin. Struct. Biol.
9
,
164
169
(
1999
).
12.
L.
Sutto
,
M.
D’Abramo
, and
F. L.
Gervasio
, “
Comparing the efficiency of biased and unbiased molecular dynamics in reconstructing the free energy landscape of Met-enkephalin
,”
J. Chem. Theory Comput.
6
,
3640
3646
(
2010
).
13.
Y.
Matsunaga
,
H.
Fujisaki
,
T.
Terada
,
T.
Furuta
,
K.
Moritsugu
, and
A.
Kidera
, “
Minimum free energy path of ligand-induced transition in adenylate kinase
,”
PLoS Comput. Biol.
8
,
e1002555
(
2012
).
14.
Y.
Mu
,
P. H.
Nguyen
, and
G.
Stock
, “
Energy landscape of a small peptide revealed by dihedral angle principal component analysis
,”
Proteins: Struct., Funct., Bioinf.
58
,
45
52
(
2005
).
15.
A.
Altis
,
P. H.
Nguyen
,
R.
Hegger
, and
G.
Stock
, “
Dihedral angle principal component analysis of molecular dynamics simulations
,”
J. Chem. Phys.
126
,
244111
(
2007
).
16.
M.
Ernst
,
F.
Sittel
, and
G.
Stock
, “
Contact- and distance-based principal component analysis of protein dynamics
,”
J. Chem. Phys.
143
,
244114
(
2015
).
17.
F.
Sittel
,
A.
Jain
, and
G.
Stock
, “
Principal component analysis of molecular dynamics: On the use of Cartesian vs. internal coordinates
,”
J. Chem. Phys.
141
,
014111
(
2014
).
18.
A.
Laio
and
M.
Parrinello
, “
Escaping free-energy minima
,”
Proc. Natl. Acad. Sci. U. S. A.
99
,
12562
12566
(
2002
).
19.
L.
Rosso
,
P.
Mináry
,
Z.
Zhu
, and
M. E.
Tuckerman
, “
On the use of the adiabatic molecular dynamics technique in the calculation of free energy profiles
,”
J. Chem. Phys.
116
,
4389
(
2002
).
20.
T.
Morishita
,
S. G.
Itoh
,
H.
Okumura
, and
M.
Mikami
, “
Free-energy calculation via mean-force dynamics using a logarithmic energy landscape
,”
Phys. Rev. E
85
,
066702
(
2012
).
21.
V.
Spiwok
,
P.
Lipovová
, and
B.
Králová
, “
Metadynamics in essential coordinates: Free energy simulation of conformational changes
,”
J. Phys. Chem. B
111
,
3073
3076
(
2007
).
22.
L.
Molgedey
and
H. G.
Schuster
, “
Separation of a mixture of independent signals using time delayed correlations
,”
Phys. Rev. Lett.
72
,
3634
3637
(
1994
).
23.
G.
Pérez-Hernández
,
F.
Paul
,
T.
Giorgino
,
G.
De Fabritiis
, and
F.
Noé
, “
Identification of slow molecular order parameters for Markov model construction
,”
J. Chem. Phys.
139
,
015102
(
2013
).
24.
H.
Takano
and
S.
Miyashita
, “
Relaxation modes in random spin systems
,”
J. Phys. Soc. Jpn.
64
,
3688
3698
(
1995
).
25.
A.
Mitsutake
,
H.
Iijima
, and
H.
Takano
, “
Relaxation mode analysis of a peptide system: Comparison with principal component analysis
,”
J. Chem. Phys.
135
,
164102
(
2011
).
26.
K.
Moritsugu
and
A.
Kidera
, “
Protein motions represented in moving normal mode coordinates
,”
J. Phys. Chem. B
108
,
3890
3898
(
2004
).
27.
R.
Car
and
M.
Parrinello
, “
Unified approach for molecular dynamics and density-functional theory
,”
Phys. Rev. Lett.
55
,
2471
(
1985
).
28.
The fictitious mass μk for vk should be small enough to adiabatically follow the variation of S. While a “try and error” procedure may be needed to set μk, a “first guess” value could be taken as the smallest mass in the physical system. In Eq. (8), we do not explicitly consider off-diagonal elements of ek due to a possible unitary transformation, which is, however, obviated by the Gram–Schmidt technique. In the CP dynamics, the SHAKE algorithm is often used to ensure the orthonormality of the electronic wave functions, where the Lagrange multipliers are introduced to evaluate the eigenvalues (ek). For a detailed explanation of the CP dynamics, see
M. C.
Payne
,
M. P.
Teter
,
D. C.
Allan
,
T. A.
Arias
, and
J. D.
Joannopoulos
, “
Iterative minimization techniques for ab initio total-energy calculations: Molecular dynamics and conjugate gradients
,”
Rev. Mod. Phys.
64
,
1045
(
1992
).
29.
P.
Focher
,
G. L.
Chiarotti
,
M.
Bernasconi
,
E.
Tosatti
, and
M.
Parrinello
, “
Structural phase transformations via first-principles simulation
,”
Europhys. Lett.
26
,
345
(
1994
).
30.
S.
Bernard
,
G.
Chiarotti
,
S.
Scandolo
, and
E.
Tosatti
, “
Decomposition and polymerization of solid carbon monoxide under pressure
,”
Phys. Rev. Lett.
81
,
2092
(
1998
).
31.
T.
Morishita
, “
Liquid–liquid phase transitions of phosphorus via constant-pressure first-principles molecular dynamics simulations
,”
Phys. Rev. Lett.
87
,
105701
(
2001
).
32.
T.
Morishita
, “
High density amorphous form and polyamorphic transformations of silicon
,”
Phys. Rev. Lett.
93
,
055503
(
2004
).
33.

Note that other algorithms to locate the extremum such as the conjugate gradient algorithm are also available.

34.
P. H.
Poole
,
F.
Sciortino
,
U.
Essmann
, and
H. E.
Stanley
, “
Phase behaviour of metastable water
,”
Nature
360
,
324
328
(
1992
);
O.
Mishima
and
H. E.
Stanley
, “
The relationship between liquid, supercooled and glassy water
,”
Nature
396
,
329
335
(
1998
).
35.
V.
Molinero
and
E. B.
Moore
, “
Water modeled as an intermediate element between carbon and silicon
,”
J. Phys. Chem. B
113
,
4008
4016
(
2009
).
36.
D. J.
Evans
,
W. G.
Hoover
,
B. H.
Failor
,
B.
Moran
, and
A. J. C.
Ladd
, “
Nonequilibrium molecular dynamics via Gauss’s principle of least constraint
,”
Phys. Rev. A
28
,
1016
1021
(
1983
).
37.
S.
Nosé
, “
Constant temperature molecular dynamics methods
,”
Prog. Theor. Phys. Suppl.
103
,
1
46
(
1991
).
38.

We also tried two additional descriptors related to the geometrical arrangement of the mW molecules in the first coordination shell, leading to TDPCA with eight descriptors. We, however, found that they were superfluous as Gs and lg (this is discussed at the end of Sec. III A 2) so that we focus on the six-descriptor analysis in Sec. III A.

39.
P. M.
Piaggi
and
M.
Parrinello
, “
Entropy based fingerprint for local crystalline order
,”
J. Chem. Phys.
147
,
114112
(
2017
) and references therein.
40.
R. M.
Puscasu
,
B. D.
Todd
,
P. J.
Daivis
, and
J. S.
Hansen
, “
Nonlocal viscosity of polymer melts approaching their glassy state
,”
J. Chem. Phys.
133
,
144907
(
2010
).
41.
M.
Matsumoto
,
S.
Saito
, and
I.
Ohmine
, “
Molecular dynamics simulation of the ice nucleation and growth process leading to water freezing
,”
Nature
416
,
409
(
2002
).
42.
T.
Morishita
, “
How does tetrahedral structure grow in liquid silicon upon supercooling?
,”
Phys. Rev. Lett.
97
,
165502
(
2006
);
[PubMed]
T.
Morishita
Structural, electronic, and vibrational properties of high-density amorphous silicon: A first-principles molecular-dynamics study
,”
J. Chem. Phys.
130
,
194709
(
2009
).
[PubMed]
43.
E. B.
Moore
,
E.
de la Llave
,
K.
Welke
,
D. A.
Scherlis
, and
V.
Molinero
, “
Freezing, melting and structure of ice in a hydrophilic nanopore
,”
Phys. Chem. Chem. Phys.
12
,
4124
4134
(
2010
).
44.
M.
Isobe
,
H.
Shimizu
, and
Y.
Hiwatari
, “
A multicanonical molecular dynamics study on a simple bead-spring model for protein folding
,”
J. Phys. Soc. Jpn.
70
,
1233
(
2001
).
45.
S.
Nosé
, “
A molecular dynamics method for simulations in the canonical ensemble
,”
Mol. Phys.
52
,
255
268
(
1984
).
46.
W. G.
Hoover
, “
Canonical dynamics: Equilibrium phase-space distributions
,”
Phys. Rev. A
31
,
1695
1697
(
1985
).
47.
H.
Grubmüller
, “
Predicting slow structural transitions in macromolecular systems: Conformational flooding
,”
Phys. Rev. E
52
,
2893
(
1995
).
48.
T.
Morishita
,
Y.
Yonezawa
, and
A. M.
Ito
, “
Free energy reconstruction from logarithmic mean-force dynamics using multiple nonequilibrium trajectories
,”
J. Chem. Theory Comput.
13
,
3106
(
2017
).
49.
T.
Morishita
and
A. M.
Itoh
, “
Travelling without dwelling: Extending the time scale accessible to molecular dynamics simulation
,”
Phys. Rev. Res.
1
,
033032
(
2019
).
50.
A.
Barducci
,
G.
Bussi
, and
M.
Parrinello
, “
Well-tempered metadynamics: A smoothly converging and tunable free-energy method
,”
Phys. Rev. Lett.
100
,
020603
(
2008
).
51.

Factor loading can be utilized when multiple PCs are invoked.

52.
P.
Das
,
M.
Moll
,
H.
Stamati
,
L. E.
Kavraki
, and
C.
Clementi
, “
Low-dimensional, free-energy landscapes of protein-folding reactions by nonlinear dimensionality reduction
,”
Proc. Natl. Acad. Sci. U. S. A.
103
,
9885
9890
(
2006
).
53.
J. R.
Errington
and
P. G.
Debenedetti
, “
Relationship between structural order and the anomalies of liquid water
,”
Nature
409
,
318
321
(
2001
).

Supplementary Material

You do not currently have access to this content.