Drug Target Commons brings Crowd Sourcing to build a Drug-Target Interaction Knowledge Base

March 06, 2018

With the costs of new therapeutic treatments under increasing scrutiny and drug development companies facing pressure to deliver new products, it has never been clearer that enormous benefit could be reaped from information sharing in the drug-target interaction space.

Dr Carl Peters | BMG LABTECH
Dr Carl Peters
PhD, Senior Applications Scientist

The authors of a recent paper in Cell Chemical Biology recognize this need and recommend that crowd-sourcing be included in the process of standardization, collection, management, curation and annotation of what is currently a highly heterogeneous data-set. Drug Target Commons (DTC) is announced by the European collaborative authors in the paper entitled: Drug Target Commons: A Community Effort to Build a Consensus Knowledge Base for Drug-Target Interactions.

A number of similar efforts have been put forth in the past. Indeed, the article identifies over 40 different resources that are designed to provide information about drug compounds, their targets and interactions. But when compared based on what the authors view as the most relevant attributes only DTC has them all. This six attributes are: free data download, bioactivity data, mutated targets (quantitative bioactivity for WT and their mutated counterparts), assay annotation, a web interface and crowd sourcing. Of the resources listed only ChEMBL could boast five of these attributes, lacking only crowd sourcing. Indeed, the authors indicate that DTC is the only resource that includes crowd sourcing as a part of the plan to build the database.


There is a good chance that you are familiar with crowd sourcing even if you are not familiar with the term. There are a number of different examples of crowd sourcing that are used to share information about traffic (Waze), create new products (Starbucks) and advertising (Greenpeace). Despite the need for moderation and feedback in crowd sourcing it serves as a way for companies to reduce cost, speed up content creation and get your target audience invested.


For the DTC, the target audience is by necessity limited in comparison to the examples above but involvement of the appropriate community will still be paramount to the success of the platform. They invite individuals to register so that they can assist in annotation of data currently available that has been added from other sources and to access the already curated data. The curated data will go through cross referencing with other databases by an approval board before an interaction is deemed suitable for indication as fully annotated.

So what kind of information is going to be included? The DTC will implement a standardized annotation that they call microBAO (micro bioassay ontology). This will include information that falls into 12 different categories: 1) Assay description, 2) Assay format, 3) Assay type, 4) Assay subtype 5) Detection technology, 6) Endpoint mode of action, 7) Inhibitor type, 8) Substrate type, 9) Substrate value, 10) Substrate unit, 11) Mutation information , 12) Assay cell line. Several of these have some very useful subcategories.  For example Assay format can be categorized as Biochemical, Cell-based, Cell-free, Organism-based, Physiochemical and Tissue-based.


Others perhaps could already use some refinement. Detection technology can be categorized as Fluorescence, Fluorescence polarization (FP), AlphaScreen, Time Resolved Fluoresecence (TRF), Time-Resolved Fluorescence Resonance Energy Transfer (TR-FRET), Label-free technology, Luminescence, Microscopy, Quantitative PCR (qPCR), Spectrophotometry and Thermal shift. Already a lot of subcategories, however, we would argue that technology like FRET and BRET likely deserve to be separated out from Fluorescence and Luminescence. Also now that there are different forms of AlphaScreen should that be acknowledged? And we know that our friends at Cisbio would say that a distinction should be made between the different TR-FRET technologies. We get it, we are a microplate reader company and detection is our business so maybe we should let it slide?


The goals of this project are quite lofty but we would judge them to be attainable. The potential benefit could be huge in projects to repurpose drugs and to create a better understanding of drugs potential adverse effects.


Of course the authors have completed some proof of principle and have already annotated over 4000 different compounds and found some interesting patterns. They present the case of BCR-ABL where they were able compare 25 different compounds and their reported effect on WT and a mutant BCR-ABL as well as Aurora kinase B (AURKB). In particular they looked into the inhibitor axitinib and noted that there are some structurally similar inhibitors that thus may be useful in repurposing. They sought to confirm this possibility using a cell based assay where viability was assessed using a luminescent readout that was read on a BMG LABTECH PHERAstar FS. They were able to confirm that axitinib was a more potent inhibitor when mutant BCR-ABL was expressed and that one of the structurally similar inhibitors, KW-2449, behaved in a similar fashion.

BMG will be watching with great interest to see whether this project is able generate enough interest within the worldwide research community. Like any crowd sourcing approach the DTC will only be able to succeed with committed participants that feel they have a stake in making sure that the goals are achieved. In the interest of bettering treatments around the world we certainly hope it is a great success!