Drug Target Commons: A Community Effort to Build a Consensus Knowledge Base for Drug-Target Interactions

Abstract

Knowledge of the full target space of bioactive substances, approved and investigational drugs as well as chemical probes, provides important insights into therapeutic potential and possible adverse effects. The existing compound-target bioactivity data resources are often incomparable due to non-standardized and heterogeneous assay types and variability in endpoint measurements. To extract higher value from the existing and future compound target-profiling data, we implemented an open-data web platform, named Drug Target Commons (DTC), which features tools for crowd-sourced compound-target bioactivity data annotation, standardization, curation, and intra-resource integration. We demonstrate the unique value of DTC with several examples related to both drug discovery and drug repurposing applications and invite researchers to join this community effort to increase the reuse and extension of compound bioactivity data.

Keywords: bioassay annotation; chemical biology; cheminformatics; community effort; crowd sourcing; data curation; drug discovery; drug repositioning; drug repurposing; open data.

Graphical abstract

Figure 1

Schematics of the DTC Platform (Open-Access Database and Web Application) The web-based platform enables the user community to take part in crowd-sourced data extraction, annotation, and curation, as well as in using and analyzing comprehensive and standardized compound-target interaction profiles. The community-driven effort aims to provide maximally high-quality and reproducible bioactivity profiles and related information that will be fed back and cross-referenced to the original data sources, therefore supplementing and enhancing the coverage and annotation of existing drug/target data resources through the crowd-sourcing initiative. Processing errors and inconsistencies in the experimental data can be minimized via open discussions, enabled by the web interface, and only the most reliable bioactivity data will be released for end users through regular updates under the Creative Commons License.

Figure 2

Bioassay Annotations Explain Heterogeneity in Bioactivity Data (A) 74 bioactivity data points for the gefitinib-EGFR drug-target pair prior to assay annotation. (B) The μBAO annotation process revealed that the major source of the variation was driven by the assay type (x axis), and further variation can be attributed to the detection technique and assay formats (colors and shapes). The low potency outliers originated from kinase assays run at very high ATP concentrations. (C) 78 bioactivity data points for the celecoxib-COX2 drug-target pair before assay annotations. (D) A clear distinction was observed in the assays performed ex vivo (human blood), compared with recombinant proteins (x axis). Further variation in the bioactivities arises from the specific target sources. See also Figures S1 and S2.

Figure 3

Compounds with Differential Potency against ABL1 (T315I) (A) A set of 25 compounds that showed potency toward phosphorylated-ABL1 (T315I), based on the current DTC database. Bubbles mark the potency class (based on half maximal inhibitory concentration [IC₅₀] in nM) of these compounds toward ABL1 (T315I), wild-type ABL1, and Aurora kinase B (AURKB), as an estimate of the potential therapeutic window. The structural similarity of the compounds is visualized as a dendrogram (constructed with the C-SPADE web tool available at http://cspade.fimm.fi; Ravikumar et al., 2017). The gray-shaded part marks candidate compounds, KW-2449 and to a lesser extent TAE-684, that are structurally similar to axitinib (an ABL1 [T315I] inhibitor), and show similar differential selectivity toward ABL1 (T315I). (B) Ba/F3 cells stably expressing BCR-ABL1 (T315I) were used for experimental validation, with compound concentrations on the x axis and the viability readout on the y axis (mean and SD errors calculated based on three or more replicates). As expected, the positive control axitinib had a higher potency toward BCR-ABL1 (T315I)-driven cells, compared with BCR-ABL1 wild-type-driven cells; similarly, KW-2449 showed a slightly higher potency toward ABL1 (T315I) compared with BCR-ABL1 wild-type. The potency of TAE-684 was actually higher toward BCR-ABL1 wild-type than toward ABL1 (T315I) in the cell-based validation, demonstrating the importance of further pre-clinical evaluations before entering the drug optimization or repurposing phases.