Attention: Data Proxy will be migrated from SWIFT to S3 storage at Friday, the 9th of May 2025 starting from 9pm CEST (my timezone). For more details, please join the rocket chat channel https://chat.ebrains.eu/channel/data-proxy-user-group


Wiki source code of Distance-Fluctuations

Last modified by emacasali on 2022/09/15 13:34
Show last authors
1 (% class="jumbotron" %)
2 (((
3 (% class="container" %)
4 (((
5 = Distance Fluctuation (DF) Analysis =
6
7 Giorgio Colombo Group (UNIPV)
8 )))
9 )))
10
11 (% class="row" %)
12 (((
13 (% class="col-xs-12 col-sm-8" %)
14 (((
15 = What DF matrices are? =
16
17 The analysis of the results of a MD simulation can performed using the Distance Fluctuation matrices (DF), based on the Coordination Propensity hypothesis:
18
19 (% style="text-align:center" %)
20 [[image:CodeCogsEqn-58.png||height="34" width="181"]]
21
22 low CP values, corresponding to low pair-distance fluctuations, highlight groups of residues that move in a mechanically coordinated way.
23
24
25 = How to use the script =
26
27 • __Requisites__
28
29 - Python 3.0 (or newer version)
30
31 - Numpy
32
33 - Scipy
34
35
36 • __Usage__
37
38 The script can analyze a MD trajectory and identify the coordinated motions between residues. It can then filter the output matrix based on the distance to identify long-range coordinated motions.
39
40 The script can work both using only C-alphas (using either a pdb or a xyz file) or the sidechains (using a pdb file).
41
42 For more information run:
43
44 {{{ python3 distance_fluctuation.py -h}}}
45
46
47 {{{optional arguments:
48 -h, --help show this help message and exit
49
50 Required arguments:
51 -ext {pdb,xyz}, --file_ext {pdb,xyz}
52 Input trajectory file iformat (options: .pdb or .xyz)
53 -n N_AA, --n_aa N_AA Number of aminoacids in each frame of the .pdb
54 trajectory
55
56 Optional arguments:
57 -i IN_NAME, --in_name IN_NAME
58 Name of the trajectory file (default: trj)
59 -s S_FRAME, --s_frame S_FRAME
60 Index of the initial frame (default = 0)
61 -e E_FRAME, --e_frame E_FRAME
62 Index of the last frame (default = end)
63 -c CUTOFF, --cutoff CUTOFF
64 Cutoff for the distance fluctuation analysis (default = 5)
65 -t TOLERANCE, --tolerance TOLERANCE
66 Tolerance for the distance fluctuation analysis (default = 0)
67 -p {all,c-a}, --pdb_type {all,c-a}
68 Type of pdb file submitted (only C-alpha, all protein atoms)
69 -l {s,seq,v,volume}, --local {s,seq,v,volume}
70 Type of local cutoff used, sequence or volume (default = seq)
71 -r RADIUS, --radius RADIUS
72 Radius of the area cutoff (default = 7A)
73 -rs RES_START, --res_start RES_START
74 Starting residue if the PDB does not start with res.num. 1
75 -ro RMS_OUT, --rms_out RMS_OUT
76 RMS distance output filename (without extension)
77 -ao AVG_OUT, --avg_out AVG_OUT
78 Average distance output filename (without extension)
79 -so SEQ_OUT, --seq_out SEQ_OUT
80 Profile sequence output filename (without extension)
81 -da DIST_AN, --dist_an DIST_AN
82 Distance analysis output filename (without extension)
83 --blocks BLOCKS File with domains borders
84
85 -rb RMS_B_OUT, --rms_b_out RMS_B_OUT
86 RMS distance blocks output filename}}}
87
88
89 )))
90
91
92 (% class="col-xs-12 col-sm-4" %)
93 (((
94 {{box title="**Contents**"}}
95 {{toc/}}
96
97
98 {{/box}}
99
100
101 )))
102 )))
103
104 • __How to read the output__
105
106 The script generate different output file.
107
108
109 A) Average Distance (avgdist) file:
110
111 The name of the file is avgdist_out_x_y_type.dat with x = start frame of the analysis, y = end frame of the analysis, type = type of analysis (CA or sidechains)
112
113 The file contains a matrix using the residue indexes as axes and the average value of the distance between the residues as the data (r1 r2  avgdist).
114
115 The distance is calculated as the average of the euclidean distance between the residues.
116
117
118 B) Distance Fluctuation (rmsdist) file:
119
120 The name of the file is rmsdist_out_x_y_type.dat with x = start frame of the analysis, y = end frame of the analysis, type = type of analysis (CA or sidechains).
121
122 The file contains a matrix using the residue indexes as axes and the distance fluctuation between the residues as the data (r1 r2 rmsdist).
123
124 The distance fluctuation is calculated for the residues that are at least 3 residues away from each other (x-2 to x+2) as follow:
125
126 1) Calculate the average euclidean distance between the residues (either CA or center of mass)
127
128 2) Calculate the average distance vector
129
130 3) Substract the distance fluctuation to the average distance
131
132 4) Calculate the power of the difference between distance fluctuation and local fluctuation
133
134 5) Filter the values of the close residues (1-x, x, 1+x)
135
136 6) Divide the obtained value for the number of frames
137
138
139 C) Profile Sequence (profile_sequence) file:
140
141 The name of the file is profile_sequence_x_y_type.dat with x = start frame of the analysis, y = end frame of the analysis, type = type of analysis (CA or sidechains).
142
143 The file is a vector containing the residue number and the local flucutation value.
144
145 The local fluctuation is calculated as the average fluctuation of the residues close to each other (that is the residues ranging from x-2 to x+2 with x = residue number).
146
147 Since the value contains the average distance fluctuation for a range of residues the output starts from residue 4 and ends at residue n-3).
148
149
150 D) Distance Analysis (distance_analysis_out) file:
151
152 The name of the file is distance_analysis_out_c_x_y_type.dat with c = cutoff value, x = start frame of the analysis, y = end frame of the analysis, type = type of analysis (CA or sidechains).
153
154 This fail contains the distance fluctuation filtered by the cutoff value (the value are kept if the distance fluctuation is smaller than the cutoff value).
155
156 The cutoff value is calculated as the sum between the nearcutoff and the tolerance value specified by the user.
157
158 The nearcutoff can be calcolated either using sequence proximity or 3d proximity and can be specified by command line using the option -l or ~-~-local (see the command line help for further details).
159
160 In the case of sequence proximity it is calculated as the average of the distance fluctuation value for residues in the range of x-2,x+2 (x = residue index)
161
162 In the case of 3d proximity is calculated as the average distance fluctuation of the residues within a certain radius from the current one (default value = 6.5 A)
163
164
165 E) Blocks Averaging: (rmsdist_out_b)
166
167 It is possible to obtain a distance fluctuation matrix average on protein domains ( or blocks).
168
169 The script requires an input file with the borders of the protein blocks defined by the user, in the form of two columns: the first with the lower limit and the second with the upper limit.
170
171
172 = References =
173
174 Morra, G.; Potestio, R.; Micheletti, C.; Colombo, G., Corresponding Functional Dynamics across the Hsp90 Chaperone Family: Insights from a Multiscale Analysis of MD Simulations, PLOS Computational Biology 8(3): e1002433. [[https:~~/~~/doi.org/10.1371/journal.pcbi.1002433>>url:https://doi.org/10.1371/journal.pcbi.1002433]]