Distinctive Behaviors of Druggable Proteins in Cellular Networks

Abstract

The interaction environment of a protein in a cellular network is important in defining the role that the protein plays in the system as a whole, and thus its potential suitability as a drug target. Despite the importance of the network environment, it is neglected during target selection for drug discovery. Here, we present the first systematic, comprehensive computational analysis of topological, community and graphical network parameters of the human interactome and identify discriminatory network patterns that strongly distinguish drug targets from the interactome as a whole. Importantly, we identify striking differences in the network behavior of targets of cancer drugs versus targets from other therapeutic areas and explore how they may relate to successful drug combinations to overcome acquired resistance to cancer drugs. We develop, computationally validate and provide the first public domain predictive algorithm for identifying druggable neighborhoods based on network parameters. We also make available full predictions for 13,345 proteins to aid target selection for drug discovery. All target predictions are available through canSAR.icr.ac.uk. Underlying data and tools are available at https://cansar.icr.ac.uk/cansar/publications/druggable_network_neighbourhoods/.

Fig 1. Enrichment and depletion of key parameters in drug targets over what can be expected at random from the interactome.

A) Graphlets and their constituent isomorphism orbits. The graph shows the graphlets and orbits, ordered by descending size and complexity, most enriched in cancer-drug targets (light blue bars). These same graphlets and orbits are either slightly depleted or not differentiated from random in targets of non-cancer drugs (dark blue). The gray line represents graphlets size and complexity (high-to-low). B) The distribution of detected community sizes and the enrichment or depletion of cancer drug targets (light blule) versus targets of drugs used to treat other diseases (dark blue). C) Box plots showing distinction of degree and google page rank; as well as the vertex modularity which distinguishes inter- versus intra-community communication of nodes. Further parameters are shown in the Supporting Information.

Fig 2. Cancer-drug targets are enriched for highly connected Graphlets.

A) Interaction network highlighting the distribution of targets of approved cancer drugs (pink); targets of approved drugs from non-cancer therapeutic areas (blue); and targets predicted to be druggable by different druggability prediction methodologies(light and dark green). Druggable proteins are spread widely across the network while targets of current approved drugs tend to cluster into few areas. B) Cumulative fraction of all drug targets covered by communities. As indicated, a small number of communities cover the majority of drug targets. C) The network communities most enriched in drug targets are listed against the fold enrichment of the number of targets found in them (compared to what can be expected at random).

Fig 3. Network profiles and interactions between targets of drug combinations.

A) Radar plots showing representative network property profiles of targets of drug combination. MEK and BRAF network property profiles are more similar to one another than the network profiles of CDKs and HMGCR. This may be related to the long-term effectiveness of the combinations of drugs targeting these proteins. B) Interactions between proteins targeted by drug combination showing high level of connectivity between targets such as EGFR, BRAF and MEK. The dotted edge indicates that no direct interaction takes place between HMGCR and the other proteins in the network.