1GMX-ANAEIG(1)                       GROMACS                      GMX-ANAEIG(1)
2
3
4

NAME

6       gmx-anaeig - Analyze eigenvectors/normal modes
7

SYNOPSIS

9          gmx anaeig [-v [<.trr/.cpt/...>]] [-v2 [<.trr/.cpt/...>]]
10                     [-f [<.xtc/.trr/...>]] [-s [<.tpr/.gro/...>]]
11                     [-n [<.ndx>]] [-eig [<.xvg>]] [-eig2 [<.xvg>]]
12                     [-comp [<.xvg>]] [-rmsf [<.xvg>]] [-proj [<.xvg>]]
13                     [-2d [<.xvg>]] [-3d [<.gro/.g96/...>]]
14                     [-filt [<.xtc/.trr/...>]] [-extr [<.xtc/.trr/...>]]
15                     [-over [<.xvg>]] [-inpr [<.xpm>]] [-b <time>] [-e <time>]
16                     [-dt <time>] [-tu <enum>] [-[no]w] [-xvg <enum>]
17                     [-first <int>] [-last <int>] [-skip <int>] [-max <real>]
18                     [-nframes <int>] [-[no]split] [-[no]entropy]
19                     [-temp <real>] [-nevskip <int>]
20

DESCRIPTION

22       gmx  anaeig analyzes eigenvectors. The eigenvectors can be of a covari‐
23       ance matrix (gmx covar) or of a Normal Modes analysis (gmx nmeig).
24
25       When a trajectory is projected on eigenvectors, all structures are fit‐
26       ted  to the structure in the eigenvector file, if present, otherwise to
27       the structure in the structure file. When no run  input  file  is  sup‐
28       plied,  periodicity  will  not be taken into account. Most analyses are
29       performed on eigenvectors -first to -last, but when -first is set to -1
30       you will be prompted for a selection.
31
32       -comp:  plot  the  vector components per atom of eigenvectors -first to
33       -last.
34
35       -rmsf: plot the RMS fluctuation per  atom  of  eigenvectors  -first  to
36       -last (requires -eig).
37
38       -proj:  calculate projections of a trajectory on eigenvectors -first to
39       -last.  The projections of a trajectory on the eigenvectors of its  co‐
40       variance  matrix  are  called principal components (pc's).  It is often
41       useful to check the cosine content of the pc's, since the pc's of  ran‐
42       dom  diffusion are cosines with the number of periods equal to half the
43       pc index.  The cosine content of the pc's can be  calculated  with  the
44       program gmx analyze.
45
46       -2d:  calculate  a 2d projection of a trajectory on eigenvectors -first
47       and -last.
48
49       -3d: calculate a 3d projection of a trajectory on the first  three  se‐
50       lected eigenvectors.
51
52       -filt: filter the trajectory to show only the motion along eigenvectors
53       -first to -last.
54
55       -extr: calculate the two extreme projections along a trajectory on  the
56       average  structure and interpolate -nframes frames between them, or set
57       your own extremes with -max. The eigenvector -first will be written un‐
58       less  -first  and  -last  have  been  set explicitly, in which case all
59       eigenvectors will be written to separate files. Chain identifiers  will
60       be added when writing a .pdb file with two or three structures (you can
61       use rasmol -nmrpdb to view such a .pdb file).
62
63   Overlap calculations between covariance analysis
64       Note: the analysis should use the same fitting structure
65
66       -over: calculate the subspace overlap of the eigenvectors in  file  -v2
67       with eigenvectors -first to -last in file -v.
68
69       -inpr:  calculate  a  matrix  of inner-products between eigenvectors in
70       files -v and -v2. All eigenvectors of both files will  be  used  unless
71       -first and -last have been set explicitly.
72
73       When  -v and -v2 are given, a single number for the overlap between the
74       covariance matrices is generated. Note that the eigenvalues are by  de‐
75       fault read from the timestamp field in the eigenvector input files, but
76       when -eig, or -eig2 are given, the corresponding eigenvalues  are  used
77       instead. The formulas are:
78
79                  difference = sqrt(tr((sqrt(M1) - sqrt(M2))^2))
80          normalized overlap = 1 - difference/sqrt(tr(M1) + tr(M2))
81               shape overlap = 1 - sqrt(tr((sqrt(M1/tr(M1)) - sqrt(M2/tr(M2)))^2))
82
83       where  M1 and M2 are the two covariance matrices and tr is the trace of
84       a matrix. The numbers are proportional to the  overlap  of  the  square
85       root  of  the  fluctuations.  The normalized overlap is the most useful
86       number, it is 1 for identical matrices and 0 when the sampled subspaces
87       are orthogonal.
88
89       When  the  -entropy  flag is given an entropy estimate will be computed
90       based on the Quasiharmonic approach and based on Schlitter's formula.
91

OPTIONS

93       Options to specify input files:
94
95       -v [<.trr/.cpt/...>] (eigenvec.trr)
96              Full precision trajectory: trr cpt tng
97
98       -v2 [<.trr/.cpt/...>] (eigenvec2.trr) (Optional)
99              Full precision trajectory: trr cpt tng
100
101       -f [<.xtc/.trr/...>] (traj.xtc) (Optional)
102              Trajectory: xtc trr cpt gro g96 pdb tng
103
104       -s [<.tpr/.gro/...>] (topol.tpr) (Optional)
105              Structure+mass(db): tpr gro g96 pdb brk ent
106
107       -n [<.ndx>] (index.ndx) (Optional)
108              Index file
109
110       -eig [<.xvg>] (eigenval.xvg) (Optional)
111              xvgr/xmgr file
112
113       -eig2 [<.xvg>] (eigenval2.xvg) (Optional)
114              xvgr/xmgr file
115
116       Options to specify output files:
117
118       -comp [<.xvg>] (eigcomp.xvg) (Optional)
119              xvgr/xmgr file
120
121       -rmsf [<.xvg>] (eigrmsf.xvg) (Optional)
122              xvgr/xmgr file
123
124       -proj [<.xvg>] (proj.xvg) (Optional)
125              xvgr/xmgr file
126
127       -2d [<.xvg>] (2dproj.xvg) (Optional)
128              xvgr/xmgr file
129
130       -3d [<.gro/.g96/...>] (3dproj.pdb) (Optional)
131              Structure file: gro g96 pdb brk ent esp
132
133       -filt [<.xtc/.trr/...>] (filtered.xtc) (Optional)
134              Trajectory: xtc trr cpt gro g96 pdb tng
135
136       -extr [<.xtc/.trr/...>] (extreme.pdb) (Optional)
137              Trajectory: xtc trr cpt gro g96 pdb tng
138
139       -over [<.xvg>] (overlap.xvg) (Optional)
140              xvgr/xmgr file
141
142       -inpr [<.xpm>] (inprod.xpm) (Optional)
143              X PixMap compatible matrix file
144
145       Other options:
146
147       -b <time> (0)
148              Time of first frame to read from trajectory (default unit ps)
149
150       -e <time> (0)
151              Time of last frame to read from trajectory (default unit ps)
152
153       -dt <time> (0)
154              Only use frame when t MOD dt = first time (default unit ps)
155
156       -tu <enum> (ps)
157              Unit for time values: fs, ps, ns, us, ms, s
158
159       -[no]w (no)
160              View output .xvg, .xpm, .eps and .pdb files
161
162       -xvg <enum> (xmgrace)
163              xvg plot formatting: xmgrace, xmgr, none
164
165       -first <int> (1)
166              First eigenvector for analysis (-1 is select)
167
168       -last <int> (-1)
169              Last eigenvector for analysis (-1 is till the last)
170
171       -skip <int> (1)
172              Only analyse every nr-th frame
173
174       -max <real> (0)
175              Maximum for projection of the eigenvector on the average  struc‐
176              ture, max=0 gives the extremes
177
178       -nframes <int> (2)
179              Number of frames for the extremes output
180
181       -[no]split (no)
182              Split eigenvector projections where time is zero
183
184       -[no]entropy (no)
185              Compute  entropy  according  to  the  Quasiharmonic  formula  or
186              Schlitter's method.
187
188       -temp <real> (298.15)
189              Temperature for entropy calculations
190
191       -nevskip <int> (6)
192              Number of eigenvalues to skip when computing the entropy due  to
193              the  quasi  harmonic  approximation.  When  you  do a rotational
194              and/or translational fit prior to the covariance  analysis,  you
195              get  3  or  6 eigenvalues that are very close to zero, and which
196              should not be taken into account when computing the entropy.
197

SEE ALSO

199       gmx(1)
200
201       More    information    about    GROMACS    is    available    at     <‐
202       http://www.gromacs.org/>.
203
205       2022, GROMACS development team
206
207
208
209
2102022.3                           Sep 02, 2022                    GMX-ANAEIG(1)
Impressum