directly involved in catalysis or substrate binding. This residue has
been proposed by a recent computational study to be coupled to
residues Val6 and Val29 [44], which have been identified as key
mutation hotspot residues that communicate by allosterically coupled dynamic networks in CypA, affecting enzyme reaction rates
[42]. In the network of hydrogen bonds, we identified as hubs the
Arg37, His92, Glu23, and Gln63 (Fig. 1c). Hydrophobic contacts
are usually crucial to maintain protein structure and stability, composing the major interactions between the residues in the protein
core, that are tightly packed and shielded to the solvent. We identified two clusters in the network of hydrophobic interactions: one
localized around the N-terminal of CypA, comprising Leu
24, Pro4, Val6, Ala26, and Ala33 and one at the other side of the
antiparallel β-sheet, comprising Val12, Leu17, Val139, Met142,
Phe145, Ile156 (Fig. 1d).
3.3 Construction
and Analysis of the IIN
PyInteraph can be used to obtain a general view of the interactions
in a protein ensemble, without considering the type of each intramolecular interaction and building a common map for their visualization and analysis. In order to do so, the different interaction
graphs calculated in the previous section are combined in a metaIntramolecular Interaction Network (IIN). In this network, an
edge between two residues is present if at least one of the interaction graphs has an edge between them. The network is by default
unweighted (i.e., all weights are set to 1.0), but weights can be
added using a knowledge-based potential implemented in PyInteraph as detailed in Sect. 3.4. The following steps allow to derive the
IIN from the calculated interaction graphs:
1. Combine interaction graphs to obtain the IIN. We used the
filter_graph tool to combine the filtered interaction graphs
described previously, by supplying the three filtered interaction
graphs using the -d option multiple times.
2. Visualization and analysis of the network. We plotted and
visualized the IIN on the reference structure, using the xPyder
plug-in for PyMOL as reported in Fig. 1e. To obtain the results
shown, we loaded the topology PDB structure in PyMOL and
used the xPyder plug-in to visualize the IIN, as detailed in the
previous paragraph. We visualized hubs, highly connected
nodes in the graph involved in three or more different interactions using the Graph analysis tools in xPyder, as explained
previously.
The IIN permits to obtain an overall description of all the most
persistent non-covalent interactions in the ensemble and their location on the protein structure. Together with the calculation of
hubs, it gives an idea of the most relevant residues in the network,
possibly important for stability and structural communication. In
the IIN, we identified several residues with high connectivity in the
Interaction Networks with PyInteraph
163
been proposed by a recent computational study to be coupled to
residues Val6 and Val29 [44], which have been identified as key
mutation hotspot residues that communicate by allosterically coupled dynamic networks in CypA, affecting enzyme reaction rates
[42]. In the network of hydrogen bonds, we identified as hubs the
Arg37, His92, Glu23, and Gln63 (Fig. 1c). Hydrophobic contacts
are usually crucial to maintain protein structure and stability, composing the major interactions between the residues in the protein
core, that are tightly packed and shielded to the solvent. We identified two clusters in the network of hydrophobic interactions: one
localized around the N-terminal of CypA, comprising Leu
24, Pro4, Val6, Ala26, and Ala33 and one at the other side of the
antiparallel β-sheet, comprising Val12, Leu17, Val139, Met142,
Phe145, Ile156 (Fig. 1d).
3.3 Construction
and Analysis of the IIN
PyInteraph can be used to obtain a general view of the interactions
in a protein ensemble, without considering the type of each intramolecular interaction and building a common map for their visualization and analysis. In order to do so, the different interaction
graphs calculated in the previous section are combined in a metaIntramolecular Interaction Network (IIN). In this network, an
edge between two residues is present if at least one of the interaction graphs has an edge between them. The network is by default
unweighted (i.e., all weights are set to 1.0), but weights can be
added using a knowledge-based potential implemented in PyInteraph as detailed in Sect. 3.4. The following steps allow to derive the
IIN from the calculated interaction graphs:
1. Combine interaction graphs to obtain the IIN. We used the
filter_graph tool to combine the filtered interaction graphs
described previously, by supplying the three filtered interaction
graphs using the -d option multiple times.
2. Visualization and analysis of the network. We plotted and
visualized the IIN on the reference structure, using the xPyder
plug-in for PyMOL as reported in Fig. 1e. To obtain the results
shown, we loaded the topology PDB structure in PyMOL and
used the xPyder plug-in to visualize the IIN, as detailed in the
previous paragraph. We visualized hubs, highly connected
nodes in the graph involved in three or more different interactions using the Graph analysis tools in xPyder, as explained
previously.
The IIN permits to obtain an overall description of all the most
persistent non-covalent interactions in the ensemble and their location on the protein structure. Together with the calculation of
hubs, it gives an idea of the most relevant residues in the network,
possibly important for stability and structural communication. In
the IIN, we identified several residues with high connectivity in the
Interaction Networks with PyInteraph
163