4. The protein or proteins in the trajectory need to be made
whole, meaning that no broken molecules should be present
because of periodic boundary conditions. It is also good practice to remove any molecule or atom that does not need to be
included in the analysis from topology and trajectory.
In the case of CypA, we modified the topology file by changing
the non-standard names assigned by GROMACS to some hydrogen atoms to the standard ones, so that charged groups of arginine
would be correctly recognized and that atomic masses would not
need to be guessed. We also added the missing chain name in the
topology pdb file (“A”) so that the output files would contain it
instead of the generic SYSTEM label, which is used when no chain
definition is supplied. In the case of CypA, we filtered the trajectory
by keeping only the protein atoms and we made sure that the
protein was whole throughout the entire trajectory using the
gmx trjconv program available in the GROMACS package.
3.2 Interaction
Networks
After preparing the system, PyInteraph can be used to calculate the
network of each intramolecular interaction type, which are hydrophobic contacts, salt bridges, and hydrogen bonds. The analysis
writes as output two main files for each interaction type. One is the
interactions file, a text file that contains the list of the non-covalent
interactions identified in the ensemble. This file can be used in the
interaction plotter PyMOL plug-in which is included in the distribution of PyInteraph (more in Note 6). The second output is a
graph adjacency matrix file containing the adjacency matrix of the
interactions with persistence values as weights, and which can be
processed by using filter_graph and graph_analysis tools and visualized on the structure using the xPyder plug-in in PyMOL (see
Note 7 for more details on the file format). The interaction networks are produced as follows:
1. Calculate non-covalent interactions between residues from the
MD trajectory. We used the pyinteraph program with options -b
to analyze salt bridges, -f to analyze hydrophobic contacts, and
-y to analyze hydrogen bonds. For each interaction type, we
collected the corresponding interaction files and the adjacency
matrix files. The files containing the topology of the system and
the trajectory were specified with -s and -t options, respectively.
We used the --ff-masses option to perform the calculation using
the definition of masses from the CHARMM27 force field (see
Note 5 for details). The output files are defined by the --sbgraph, --hb-graph, --hc-graph options for the graph adjacency
matrix files of salt bridges, hydrogen bonds, and hydrophobic
clusters, respectively. We calculated the hydrogen bonds considering only the donor-acceptor atoms located on the side
chains of the residues using the option --hb-class sc-sc, in order
to exclude hydrogen bonds involved in the formation of the
160
Matteo Lambrughi et al.
whole, meaning that no broken molecules should be present
because of periodic boundary conditions. It is also good practice to remove any molecule or atom that does not need to be
included in the analysis from topology and trajectory.
In the case of CypA, we modified the topology file by changing
the non-standard names assigned by GROMACS to some hydrogen atoms to the standard ones, so that charged groups of arginine
would be correctly recognized and that atomic masses would not
need to be guessed. We also added the missing chain name in the
topology pdb file (“A”) so that the output files would contain it
instead of the generic SYSTEM label, which is used when no chain
definition is supplied. In the case of CypA, we filtered the trajectory
by keeping only the protein atoms and we made sure that the
protein was whole throughout the entire trajectory using the
gmx trjconv program available in the GROMACS package.
3.2 Interaction
Networks
After preparing the system, PyInteraph can be used to calculate the
network of each intramolecular interaction type, which are hydrophobic contacts, salt bridges, and hydrogen bonds. The analysis
writes as output two main files for each interaction type. One is the
interactions file, a text file that contains the list of the non-covalent
interactions identified in the ensemble. This file can be used in the
interaction plotter PyMOL plug-in which is included in the distribution of PyInteraph (more in Note 6). The second output is a
graph adjacency matrix file containing the adjacency matrix of the
interactions with persistence values as weights, and which can be
processed by using filter_graph and graph_analysis tools and visualized on the structure using the xPyder plug-in in PyMOL (see
Note 7 for more details on the file format). The interaction networks are produced as follows:
1. Calculate non-covalent interactions between residues from the
MD trajectory. We used the pyinteraph program with options -b
to analyze salt bridges, -f to analyze hydrophobic contacts, and
-y to analyze hydrogen bonds. For each interaction type, we
collected the corresponding interaction files and the adjacency
matrix files. The files containing the topology of the system and
the trajectory were specified with -s and -t options, respectively.
We used the --ff-masses option to perform the calculation using
the definition of masses from the CHARMM27 force field (see
Note 5 for details). The output files are defined by the --sbgraph, --hb-graph, --hc-graph options for the graph adjacency
matrix files of salt bridges, hydrogen bonds, and hydrophobic
clusters, respectively. We calculated the hydrogen bonds considering only the donor-acceptor atoms located on the side
chains of the residues using the option --hb-class sc-sc, in order
to exclude hydrogen bonds involved in the formation of the
160
Matteo Lambrughi et al.