GSK brings scientific and technical expertise to the Challenge.

Winners of the Discovery Fast Track Challenge will be given access to our world-class knowledge in drug discovery and access to GSK’s proprietary DNA-encoded compound library comprised of over a trillion unique compounds tagged with specific DNA sequences. Using our expertise in drug discovery, we will provide the chemical structures of high-quality chemical probes from successful screens according to the four possible research outcomes.


Using the purified tagged protein reagents provided by principal investigators, scientists at GSK will optimize assay conditions for various affinity selection protocols to identify novel compounds that bind to the target. In the case of enzymes, for example, active site competitive inhibitors and allosteric modulators may be identified depending on the selection conditions employed. Investigators or their delegates will be offered the opportunity to work side by side with GSK scientists and engage in hands-on activities with affinity selection protocols and interpreting screening results.

DNA-Encoded Compound Library Technology (ELT)

ELT is an affinity-based selection methodology for hit identification. This technology is based on the synthesis and screening of a highly diverse collection of small molecules tagged with specific DNA sequences used as unique barcodes. Our current collection spans 94 different libraries that, when combined, contain greater than 1 trillion unique compounds, each with its own unique DNA barcode. Our libraries exploit split and pool combinatorial chemistry to produce this diversity, while the DNA tags obviate the traditional challenge of library deconvolution.

diagram of assembly and encoding of on-DNA compounds
Figure 1. Assembly and encoding of on-DNA compounds. All libraries start from our ‘head-piece’, a covalently-linked DNA duplex with a handle for organic chemistry. Iterative rounds of chemistry and DNA ligation assemble and encode the small molecules.

In an affinity selection experiment, libraries of DNA-tagged small molecules containing over one trillion compounds are exposed to a drug target. Following washing protocols to remove weak binders, the target protein is denatured to release the tighter binding compounds or hits. The DNA tags associated with the hits are then amplified by PCR and sequenced. The DNA sequences reveal through informatic translation the structure activity relationships of the hits. Finally, attractive chemotypes are synthesized without a DNA tag and further evaluated in binding and functional biological assays to assess hit quality.

diagram of ELT hit identification process
Figure 2. The ELT hit identification process. Targets are validated before selection to confirm capture and activity under selection conditions. DNA-encoded ligands are isolated by affinity selection, sequenced and deconvoluted back to chemical structures. In order to confirm binding and activity, the identified structures need to be resynthesized in the absence of the unique DNA tag.

Selections can be run in multiple conditions in parallel. The process requires 1 milligram of protein (<100 microgram per selection and up to 10 selection conditions are employed) with an affinity tag (polyhistidine, FLAG and biotin tags are preferred). Protein quality is critical: proteins of high purity (>90%) with no aggregation significantly increase ELT success rates. The ELT has been applied successfully to a range of enzyme, receptor and protein-protein interaction targets, including target proteins isolated directly from patients.¹

Hit Qualification Support

After completing the affinity selection and DNA sequencing of the hits, GSK will resynthesize exemplars without the DNA tag, and the principal investigator will confirm and characterize the functional activity of the hits. Selecting the optimal tool compound(s) from the hundreds or thousands of raw screening hits typically requires using a range of assays (e.g., biochemical, biophysical, cell-based), along with computational chemistry approaches to identify emergent structure-activity relationships. To accomplish this, GSK operates a fully integrated environment and takes a holistic computational and experimental approach. We provide both the relevant measurements and the interpretational guidance, ensuring the selected compounds possess the best chemical properties and interact with the target via a desired modality.