Synthesis of DNA Encoded Libraries for Drug Discovery

Traditionally high throughput screening of large compound collections has been the starting point for many drug discovery campaigns. This method is time-consuming and can be resource intensive if robotic setups are not utilized. Furthermore, access to large compound collections is limited, and usually only large pharmaceutical companies have access to the vast number of compounds required for comprehensive screening. One solution to the challenges associated with high-throughput screening is DELs, which offer a more efficient and cost-effective approach to screening millions of compounds in parallel. 

DELs have the unique feature of combining the power of chemistry with DNA to create libraries of compounds that can easily be screened and decoded rapidly. In DELs the compounds are linked covalently to short DNA sequences, which serve as their identifiers. This allows for fast and accurate identification of hits during screening by decoding these using next-generation sequencing (NGS). 

Multiple techniques exist to construct DELs, and novel approaches also appear in literature today. There are, however, a set of methods that have proven powerful and are used today to construct DELs both in academia as well as in pharmaceutical companies. The most prominent methods are described below.

Split-and-Pool Synthesis is the most widely used technique for constructing DELs (Figure 1A). Here, the library is constructed stepwise by subsequent chemical reactions and DNA encoding. Following each encoding step, the library is pooled together and mixed before it is split into new fractions for the subsequent chemical step. If N building blocks are utilized in each round (R) of split-and-pool synthesis the library will consist of NR library members. Three rounds of synthesis with 100 building blocks in each round will result in a DEL containing 106 different molecules. If one additional round is added the library will have 108 different molecules. It might seem compelling to add more rounds of synthesis, but practically once the synthesis exceeds 3-4 rounds the molecules identified will have a high molecular weight which falls outside of the drug-like chemical space. 

The major pitfall of DEL synthesis by Split-and-Pool is the risk of truncated products. As the encoding (e.g. by ligation) of DNA happens independently of the chemical reaction each individual molecule will carry the DNA-tag even if the chemical reaction has not taken place. This puts limits on which chemical reaction can take place as they need to be high yielding and not generate side products. 

Encoded Self-Assembling Chemical (ESAC) Libraries rely on single stranded DNA to form duplexes when paired with a complementary strand. This allows for the construction of sub-libraries containing a constant region which has a complementary hybridization strand, which forms duplexes with the other sub-libraries. If each sub-library contains X and Y members respectively, the total number of library members will be X×Y. ESAC libraries can be utilized in many ways allowing for a flexible screening platform:

• Two sub-libraries can be assembled allowing for identification of de novo bidentate molecules. Though this might seem appealing, the initial follow-up work needs to merge these individual ligands together, which can be challenging.
• A non-coding oligonucleotide can be used to make a sub-library double stranded. This will give display of one of the sub-libraries which will resemble a classic DEL.
• Pairing a known binder with a sub-library allows for the maturation of a hit compound.

Where Split-and-Pool synthesis is described as DNA-recorded, utilizing the YoctoReactor can be classified as DNA-templated DEL synthesis as the DNA sequence and formation of the 3D template drive library construction. Other methods of DNA-templated DEL synthesis exist, e.g. by Halpin and Harbury, who developed hybridization columns which can capture DNA onto Sepharose resin on which the chemical reaction can be performed (Halpin and Harbury 2004). A similar technique was developed by Li and co-workers, who introduced a universal DNA template, which can hybridize with non-specific reagent DNAs containing different sequences (Li 2013). In this way, one DNA string can be used to template the entire synthesis.

DNA-Encoded Libraries (DELs) have transformed the landscape of early-stage drug discovery by enabling the rapid identification of small-molecule binders across a broad range of biological targets. Beyond their utility as a screening platform, DELs have been integrated into various stages of hit identification and optimization, with growing relevance in both academic and industrial settings.

One of the key advantages of DEL technology lies in its capacity to screen billions of compounds simultaneously using minimal amounts of target protein. This has made DELs especially attractive for targets that are challenging to access using traditional high-throughput screening (HTS) methods. These include:

• Protein–protein interactions (PPIs): DELs have been successfully applied to identify small molecules that modulate or inhibit PPIs, a target class traditionally considered undruggable.
• Difficult or low-abundance targets: DELs can be used even when target protein is available only in small quantities, such as membrane proteins, nuclear receptors, and multi-subunit complexes.
• Fragment-based approaches: DELs can be used to explore vast chemical space with small, fragment-like building blocks, which can be further optimized post-screening.

Once potential binders are identified through DEL selection and decoding, off-DNA synthesis of hits enables their validation in orthogonal biochemical or biophysical assays. Many companies have incorporated DEL-derived hits into their lead optimization workflows, where these compounds undergo structure–activity relationship (SAR) exploration and pharmacological profiling.

DELs have also played a role in:

• Allosteric modulator discovery, where hits bind to non-canonical sites
• Covalent inhibitor development, through libraries containing reactive electrophiles

Structure-based drug design, especially when DEL hits are co-crystallized with targets

DELs have been successfully used across various target classes, including:

• Kinases
• G-protein-coupled receptors (GPCRs)
• Proteases
• Epigenetic regulators
• RNA-binding proteins

This versatility makes DELs a powerful complementary approach to traditional screening platforms, especially when combined with structural biology or AI-driven hit expansion.

DNA-Encoded Libraries (DELs) represent a transformative advancement in early-stage drug discovery, offering an efficient and scalable means to explore vast chemical space. Through diverse synthesis strategies such as split-and-pool and proximity-driven approaches like Vipergen’s YoctoReactor, DELs enable the rapid identification of small-molecule binders against a wide range of biological targets. Each method offers unique advantages—split-and-pool synthesis provides unparalleled diversity, while proximity-driven methods offer higher fidelity and reduced truncation artifacts. Together, these complementary technologies allow researchers to tailor DEL campaigns to specific targets, chemical constraints, or discovery goals. As DEL platforms continue to evolve, they are expected to play an increasingly central role in hit discovery, target validation, and the acceleration of lead optimization efforts.

• Brenner, S. and Lerner, R. A. (1992), Encoded combinatorial chemistry. Proc. Natl. Acad. Sci. USA, 89, 5381-5383. doi.org/10.1073/pnas.89.12.5381
• Gartner, Z. J. et. al. (2004), DNA-Templated Organic Synthesis and Selection of a Library of Macrocycles. Science, 305, 1601-1605. doi.org/10.1126/science.1102629
• Halpin, D. R. and Harbury, R. B. (2004), DNA Display I. Sequence-Encoded Routing of DNA Populations, PLoS Biology, 2 (7), 1015-1021. doi.org/10.1371/journal.pbio.0020173
• Halpin, D. R. and Harbury, R. B. (2004), DNA Display II. Genetic Manipulation of Combinatorial Chemistry Libraries for Small-Molecule Evolution, PLoS Biology, 2 (7), 1022-1030. doi.org/10.1371/journal.pbio.0020174
• Hansen, M. H. et. al. (2009), A Yoctoliter-Scale DNA Reactor for Small-Molecule Evolution, J. Am. Chem. Soc., 131 (3), 1322-1327. doi.org/10.1021/ja808558a
• Li, Y., et. al. (2013), Multistep DNA-Templated Synthesis Using a Universal Template. J. Am. Chem. Soc., 135, 17727-17730. doi.org/10.1021/ja409936r
• Melkko, S. et. al. (2004), Encoded Self Assembling Chemical Libraries. Nat. Biotechnol., 22, 568-574. doi.org/10.1038/nbt961
• Nielsen, J., Brenner, S., Janda, K. D. (1993), Synthetic methods for the implementation of encoded combinatorial chemistry. J. Am. Chem. Soc., 115, 9812-9813. doi.org /10.1021/ja00074a063