On June 14, 2021, the article "DNA-encoded chemical libraries: a comprehensive review with successful stories and future challenges" provides a comprehensive examination of the groundbreaking review "DNA-encoded chemical libraries: a comprehensive review with successful stories and future challenges" by Gironda-Martinez et al., published in ACS Pharmacology & Translational Science. This influential analysis traces the remarkable evolution of DNA-encoded library (DEL) technology from its conceptual origins in Brenner and Lerner's 1992 pioneering work to the current cutting-edge integration of machine learning and artificial intelligence. The review demonstrates how DELs have transformed drug discovery by enabling the creation and screening of ultra-large compound collections that far exceed traditional high-throughput screening capabilities, particularly through innovative encoding strategies and screening methodologies. These advancements are perfectly aligned with CD BioGlyco's specialized DNA-encoded Glycan Library (DEGL) Services, which leverage these technological breakthroughs to advance glycan-protein interaction studies and therapeutic development.

Historical Development and Technological Evolution

The concept of DNA-encoded libraries was first proposed by Brenner and Lerner in 1992, who envisioned synthetic chemical entities on beads linked to DNA fragments serving as identification barcodes. As highlighted in the review, it took approximately 15 years for this pioneering idea to gain significant traction, with three key methodologies emerging in 2004 that featured direct coupling of chemical matter to double-stranded DNA fragments without solid supports. These included DNA-templated synthesis developed by Liu and colleagues, encoded self-assembling chemical (ESAC) libraries by Neri and coworkers, and DNA-routing technology from Harbury's group.

Fig.1 Schematic of encoding strategies for DNA-recorded chemical libraries. (Gironda-Martínez, et al., 2021)

CD BioGlyco has advanced these foundational technologies to create specialized platforms for glycan research. Our DEGL services incorporate the most effective encoding strategies, including both DNA-recorded and DNA-templated synthesis approaches, optimized specifically for carbohydrate chemistry and glycan-protein interaction studies.

Library Encoding Strategies and Design Principles

The review article expertly categorizes DEL encoding strategies into single-pharmacophore and dual-pharmacophore libraries. Single-pharmacophore libraries typically employ DNA-recorded synthesis using split-and-pool procedures, where libraries are built through sequential chemical transformations, each encoded by the addition of DNA fragments that uniquely identify them. Dual-pharmacophore libraries, particularly ESAC libraries, enable the display of pairs of building blocks that can synergistically interact with target proteins.

Fig. 2 Encoding strategy for dual-pharmacophore DNA-encoded libraries via the ESAC approach. (Gironda-Martínez, et al., 2021)

At CD BioGlyco, we have developed sophisticated DEGL Design Capabilities that incorporate both approaches. Our Biology-driven and AI-based DEGL Design Services leverage a deep understanding of glycan structures and their biological significance to create diverse, high-quality library collections. The platform includes DNA-encoded Single-pharmacophore, Dual-pharmacophore, and Trio-pharmacophore glycan libraries, as well as specialized collections targeting specific biological applications.

Screening Methodologies and Applications

DEL screening methodologies have evolved significantly beyond traditional solid-phase affinity selections. While magnetic bead-based capture remains popular, researchers have developed sophisticated solution-phase screening techniques that preserve protein native conformation and function.

interaction-dependent PCR (IDPCR) represents a particularly innovative approach where target proteins are conjugated with oligonucleotides complementary to DEL tags. This methodology stabilizes ligand-protein interactions and enables selective PCR amplification of binding molecules. DNA-programmed photoaffinity labeling (DPAL) represents another significant advancement, allowing covalent capture of binding interactions without prior protein immobilization.

Fig.3 Schematic representation of three evolving hi-EDCCL methodologies. (Gironda-Martínez, et al., 2021)

Cell-based DEL screening has expanded the technology's applicability to membrane proteins and targets in their native biological contexts. Recent methodologies now enable screening against proteins expressed on live cells, opening new possibilities for targeting complex biological systems. The practical applications of DEL technology have yielded numerous success stories, with several DEL-derived compounds advancing to clinical trials. These include inhibitors of receptor-interacting protein 1 (RIP1) kinase, soluble epoxide hydrolase (sEH), and autotaxin (ENPP2). The technology has proven particularly valuable for targeting protein-protein interactions and other challenging target classes that have historically been difficult to drug with small molecules.

The practical applications of DEL technology have yielded numerous success stories, with several DEL-derived compounds advancing to clinical trials. These include inhibitors of receptor-interacting protein 1 (RIP1) kinase, soluble epoxide hydrolase (sEH), and autotaxin (ENPP2). The technology has proven particularly valuable for targeting protein-protein interactions and other challenging target classes that have historically been difficult to drug with small molecules.

Current Challenges and Future Directions

Despite significant progress, DEL technology faces several challenges. False positive and false negative rates remain concerns, particularly with libraries exceeding 10^8 members. Statistical methods for analyzing sequencing data continue to evolve, with recent approaches incorporating machine learning to distinguish true binders from background noise.

The integration of artificial intelligence and machine learning represents perhaps the most promising future direction for DEL technology. These approaches can help design better libraries, predict compound properties, and analyze screening results more effectively. As DEL datasets grow in size and complexity, AI-driven analysis will become increasingly essential for extracting meaningful insights.

Chemical space coverage remains another challenge, though ongoing development of DNA-compatible reactions continues to expand structural diversity. Transition-metal-promoted reactions, radical-based transformations, and novel heterocycle formations are particularly active areas of research.