network and possible communication hubs: Val12, Glu23, Leu24,
Lys31, Arg37, Glu43, Gln63, His92, Val139, Met142, and Ile156.
The central role in the network of contacts of the Gln63 is particularly interesting since it interacts directly with the substrate through
hydrogen bonds, and both NMR experiments and computational
investigations identified this residue as involved in the major
conformational processes of CypA [40, 49]. It should be noted
that the IIN contains all the analyzed interactions in the network,
without taking into account nonspecific contacts that can still take
part in communication pathways in the network. For this reason,
we also considered a more generic analysis, cmPSN, which
takes into account any possible contact between residues (see Sect.
3.5).
3.4 Energy
Interaction Network
As outlined in the introduction, PyInteraph can be used to calculate
average interaction knowledge-based potential pseudo-energies based on sets of four distances between pairs of residues. This
is especially useful to have an idea of how favorable the interactions
identified in the IIN or in the cmPSN are, especially for the IIN as it
has no associated weights by default. Given the definition of the
knowledge-based potential, negative values account for favored
interactions while positive values account for unfavored interactions. The weighted IIN has been obtained as follows:
1. Calculate the interaction energy potential for all possible pairs
of residues. This is performed similarly as the other analyses,
using the pyinteraph program with the -p option. Lists of
interaction energies per pair of residues are written in the file
specified by --kbp-dat while the same information in adjacency
matrix form is saved with --kbp-graph. The calculated energy
network can be used as-is or it can be used to weight any of the
networks we calculate.
2. Assign weights to the IIN. This is done using the filter_graph
executable, with option -d for the IIN adjacency matrix file and
option -w for the knowledge-based potential graph.
3. Visualize the weighted IIN. We plotted the obtained weighted
graph using xPyder, as detailed above for the other interaction
graphs. See Note 9 for more details on network analysis on this
type of network.
The obtained energy network can be used to understand which
interactions are most favored among those identified and is shown
in Fig. 2. All except one of the identified interactions are found to
be somehow favorable when scored by the knowledge-based potential, however, with different magnitudes. As expected, there is a
degree of correlation between the most stable interactions and their
strengths, although some highly persistent interactions are not
considered to be strong when scored with the potential. This is
164
Matteo Lambrughi et al.
Lys31, Arg37, Glu43, Gln63, His92, Val139, Met142, and Ile156.
The central role in the network of contacts of the Gln63 is particularly interesting since it interacts directly with the substrate through
hydrogen bonds, and both NMR experiments and computational
investigations identified this residue as involved in the major
conformational processes of CypA [40, 49]. It should be noted
that the IIN contains all the analyzed interactions in the network,
without taking into account nonspecific contacts that can still take
part in communication pathways in the network. For this reason,
we also considered a more generic analysis, cmPSN, which
takes into account any possible contact between residues (see Sect.
3.5).
3.4 Energy
Interaction Network
As outlined in the introduction, PyInteraph can be used to calculate
average interaction knowledge-based potential pseudo-energies based on sets of four distances between pairs of residues. This
is especially useful to have an idea of how favorable the interactions
identified in the IIN or in the cmPSN are, especially for the IIN as it
has no associated weights by default. Given the definition of the
knowledge-based potential, negative values account for favored
interactions while positive values account for unfavored interactions. The weighted IIN has been obtained as follows:
1. Calculate the interaction energy potential for all possible pairs
of residues. This is performed similarly as the other analyses,
using the pyinteraph program with the -p option. Lists of
interaction energies per pair of residues are written in the file
specified by --kbp-dat while the same information in adjacency
matrix form is saved with --kbp-graph. The calculated energy
network can be used as-is or it can be used to weight any of the
networks we calculate.
2. Assign weights to the IIN. This is done using the filter_graph
executable, with option -d for the IIN adjacency matrix file and
option -w for the knowledge-based potential graph.
3. Visualize the weighted IIN. We plotted the obtained weighted
graph using xPyder, as detailed above for the other interaction
graphs. See Note 9 for more details on network analysis on this
type of network.
The obtained energy network can be used to understand which
interactions are most favored among those identified and is shown
in Fig. 2. All except one of the identified interactions are found to
be somehow favorable when scored by the knowledge-based potential, however, with different magnitudes. As expected, there is a
degree of correlation between the most stable interactions and their
strengths, although some highly persistent interactions are not
considered to be strong when scored with the potential. This is
164
Matteo Lambrughi et al.