Overview

Cyclic peptidomimetics (CPM) have attracted growing attention in drug discovery because they combine the developability of small molecules with the target-recognition capability of larger biomolecules. Yet their complex macrocyclic topologies, noncanonical amino acids, and diverse cross-linking chemistries continue to challenge conventional AI-based molecular modeling methods.

Recently, the Computational Chemistry team at HitGen introduced CycWeave, a token-free dual-view coarse-grained graph neural framework designed for complex modular molecular systems. By adopting a representation strategy that better matches the modular nature of CPMs and DNA-encoded library (DEL) compounds, CycWeave demonstrated robust and competitive performance in both CPM membrane permeability prediction and large-scale DEL enrichment modeling. (Preprint available on ChemRxiv, https://chemrxiv.org/doi/full/10.26434/chemrxiv.15001512/v1)

01 Challenges in Current Computational Modeling

In AI-driven drug discovery, computational modeling of CPMs and structurally complex DEL compounds faces two major limitations:

1. Atom-level graphs often fail to capture global topology

Conventional graph neural networks (GNNs) primarily focus on local atoms and bonds, but often struggle to effectively represent the higher-order topological organization characteristic of cyclic peptide-like systems, such as scaffold architecture, branch placement, and connection patterns.

2. Vocabulary-dependent token models have limited generalization

Many existing peptide or fragment-based modeling methods rely on predefined vocabularies or tokenization schemes. In realistic CPM-oriented DEL settings, however, noncanonical monomers and open-ended chemical modifications are common. As a result, such methods can suffer from out-of-vocabulary limitations and reduced generalizability in open chemical space.

Figure 1. Summary of existing molecular modeling strategies for CPM

02 Core Design Logic of CycWeave

To address these challenges, CycWeave introduces a new representation framework specifically designed for structurally complex and modular molecules.

1. Dual-view graph architecture

CycWeave represents each molecule simultaneously as an atom-level graph and a fragment-level coarse-grained graph. The atom-level view captures local chemical environments, while the coarse-grained view explicitly preserves modular structure by decomposing molecules into scaffold, branch, and connection-level components, including key chemical relations such as amide linkages, ring connection sites, and disulfide bonds. The two views are coupled and fused within a unified neural architecture, enabling coordinated modeling of both local detail and global topology.

2. Token-free continuous fragment embeddings

A central innovation of CycWeave is its token-free design. Instead of mapping fragments into discrete symbolic tokens, the framework uses continuous ECFP-based fragment embeddings to initialize coarse-grained nodes. This avoids dependence on a fixed vocabulary and enables the model to generalize more naturally to novel noncanonical monomers and open-ended chemical modifications.

3. Support for self-supervised pretraining

CycWeave also supports a self-supervised pretraining–fine-tuning paradigm. Through a masked fragment recovery task, the model learns to reconstruct original continuous fragment fingerprints from surrounding structural context. This allows CycWeave to learn transferable structural priors from large unlabeled DEL-related CPM chemical spaces and improves its applicability to downstream tasks with limited labeled data.

Figure 2. Schematic overview of the token-free coarse-grained dual-view framework of CycWeave.

03 Application Validation: Developability Assessment and DEL Screening Modeling

The research team systematically evaluated CycWeave in two practically important application scenarios.

1. CPM membrane permeability prediction

Membrane permeability is jointly influenced by local physicochemical features and higher-order structural organization. On public benchmark datasets including PAMPA, Caco-2, MDCK, and RRCK, CycWeave achieved the strongest overall performance on the major benchmarks after pretraining and fine-tuning. Notably, it reached an R² of 0.728 in Caco-2 and 0.701 on the aggregated dataset, outperforming representative intermediate-granularity baselines such as PepLand and PeptideCLM. These results support the value of token-free dual-view representation for developability-related property prediction.

2. DEL enrichment modeling against TfR1

The team further applied CycWeave to DEL enrichment modeling against transferrin receptor 1 (TfR1), a biologically and translationally relevant target in drug delivery research. Because DEL enrichment signals are count-derived and typically overdispersed, the model used a negative binomial negative log-likelihood loss rather than a simple mean squared error objective. Under 10-fold scaffold-split evaluation, CycWeave outperformed both the general-purpose graph learning baseline Chemprop and the classical ECFP-MLP baseline. It achieved R² = 0.596, AUC-ROC = 0.962, and AP = 0.764, demonstrating strong regression fit as well as effective prioritization of enriched compounds under class imbalance.

In addition, latent-space visualization using t-SNE showed that enriched DEL compounds were organized into multiple separated yet internally compact clusters, suggesting that CycWeave not only improves predictive performance but may also help reveal distinct latent chemotypes or scaffold series for downstream hit triaging and series analysis.

04 Summary and Outlook

The results of CycWeave suggest that, for complex modular molecular systems such as cyclic peptidomimetics and DEL compounds, chemically meaningful coarse-grained decomposition combined with a token-free open representation can substantially improve computational modeling performance.

As a unified molecular representation backbone, CycWeave is expected to support not only CPM property prediction, but also a broader range of AI-for-chemistry applications, including DEL activity modeling, selectivity analysis, pharmacokinetic property prediction, and multi-objective molecular optimization.

Marissa D Dolorfino, Daniel Santos Perez, Yao Fu, Shu-Hang Lin, Sean McCarty, Matthew James O'Meara, Terra Sztain

bioRxiv - Biophysics

DOI: 10.64898/2026.04.18.719394

Abstract

DNA-encoded libraries (DELs) enable ultra-large screening of billions of molecules simultaneously. However, various limitations of DELs have prompted interest in training machine learning (ML) models on these large datasets to extrapolate predictions to non-DEL compounds. A recent NeurIPS competition revealed that even top performing ML models trained on DEL data failed at generalizing to out-of-distribution (OOD) chemical space. We investigated whether integrating structural modeling could bridge this generalization gap. We systematically assessed state-of-the-art ML, docking, and co-folding methods with three biologically diverse protein targets screened against libraries containing multiple DEL synthesis formats, and show that while ML excels in-distribution, the optimal approach for OOD hit discrimination performance is both target and ligand dependent. We conclude that, regardless of performance reported in aggregated benchmarks, rigorous, system-dependent pilot testing is critical for reliable virtual screening predictions. We provide these workflows and analysis tools in an open-source package: DEL-iver.

Lulu Wen,Qingqing Zhang,Zhiqiang Duan,Rui Jin,Xiaojie Lu

ChemMedChem

DOI: 10.1002/cmdc.202501032

Abstract

DNA-encoded library (DEL) technology has emerged as a powerful tool to accelerate drug discovery, and its application has expanded to challenging targets such as E3 ubiquitin ligases, whose ligands are essential for the development of targeted therapies, including proteolysis-targeting chimeras (PROTACs). In this review, we summarize recent advances in the use of DELs for the discovery of small-molecule non-covalent E3 ligase ligands and discuss their advantages in hit-to-lead optimization and the design of targeted protein degradation systems. Furthermore, we highlight the potential and application basis of covalent DELs and DNA-encoded cyclic peptide libraries, which together outline promising future directions for DEL-based discovery of E3 ligase ligands. Emerging DEL-based strategies for the direct discovery and optimization of TPD molecules are also discussed.

Ruolan Zhou, Jiajia Wang, Xiang Li, Yulong An

ACS Medicinal Chemistry Letters

DOI: 10.1021/acsmedchemlett.6c00113

Abstract

DNA-encoded library (DEL) technology has emerged as a transformative platform for the discovery of bioactive small molecules against challenging therapeutic targets including G protein-coupled receptors (GPCRs). As a clinically pivotal class of membrane-bound targets, GPCRs pose inherent challenges in the discovery of novel ligands. This Microperspective highlights recent methodological advances (2015–2026) that enable DEL selections against GPCRs, thereby facilitating the identification of diverse ligand modalities, including agonists, antagonists, allosteric modulators, and biased ligands. Furthermore, we discuss current challenges and future directions in the application of DEL technology to GPCR drug discovery, with a specific emphasis on opportunities in receptor stabilization, selection strategy design, and computational method development.

J. Miguel Mata, Jingming Liu, Sean M. McKenna, Edith van der Nol, Marije Havermans, Ruud Delwel, Mike Filius, Chirlmin Joo, Maura Vallaro, Giulia Caron, Sebastian J. Pomplun

Nature Communications

DOI: 10.1038/s41467-026-71641-3

Abstract

Synthetic macrocycles offer exceptional potential as therapeutics. However, most high-throughput discovery platforms rely on genetically encoded libraries of large peptide macrocycles, which typically are not optimized for drug like properties. Fully synthetic libraries offer greater flexibility in accessing broader chemical space. Leveraging recent advances in mass spectrometry based library techniques, here we report CycloSEL (Cyclic Self-Encoded Libraries), an end-to-end workflow, that screens synthetic macrocycle libraries enriched in drug-like ‘beyond rule of five’ features. The workflow relies on affinity selections and hit identification by tandem mass spectrometry, eliminating the need for genetic barcodes. We construct a 16 million-member library and validate the approach against the oncology target carbonic anhydrase IX, achieving robust enrichment and accurate identification of true binders. Applying CycloSEL to the acute myeloid leukemia target WD repeat-containing protein 5 (WDR5) yields a macrocycle with subnamolar affinity, and potent inhibition of the WDR5–Mixed-Lineage Leukemia 1 (MLL1) interaction. Subsequent modifications produce a chameleonic macrocycle with passive membrane permeability, serum stability, and anti-proliferative activity in leukemia cells. Together, these results demonstrate that CycloSEL enables discovery of drug-like macrocycles from fully synthetic libraries for intracellular targets.

As the field of traditional DNA-Encoded Library (DEL) chemistry reaches maturity, expectations for library quality have become increasingly exacting. Beyond conventional optimizations focused on purification protocols and reaction yields, a growing number of research groups have pioneered solid-phase synthesis strategies to enhance peptide library purity. Here, we present and discuss the key insights gleaned from three recent publications on solid-phase DEL derivatives.

Shiyu Chen et al. pioneered a solid-phase purification strategy for DNA-encoded peptide libraries (PDELs) by engineering a modified Fmoc (mFmoc) protecting group equipped with a terminal azido1. This design enables the specific immobilization of desired peptide intermediates onto alkyne-functionalized controlled pore glass (CPG) beads via copper-free click chemistry following each coupling step (Fig. 1). After rigorous washing to eliminate unreacted building blocks and truncated byproducts, the pure products are released through standard Fmoc deprotection. This "capture-and-release" cycle successfully facilitated the construction of the longest reported five-round PDEL with purity exceeding 95%, effectively breaking the conventional four-round synthesis barrier. (Solid DEL-1)

Fig. 1: Iterative cycles of generating a purified DNA-encoded peptide library with mFmoc-protected amino acids. The desired DNA-encoded peptide is isolated after immobilization and purification (Solid DEL-1).

The group of Jörg Scheuermann developed another dual-linker solid-phase synthesis strategy on magnetic beads to achieve "self-purifying" release of DELs (Fig. 2)2. This solid-phase platform not only facilitates the synthesis of high-purity five-cycle desired peptide compounds but also significantly expands the compatible reaction scope to include water-free conditions, enabling transformations previously inaccessible in traditional aqueous DEL synthesis, such as the SnAP cyclization reaction and acid-mediated Boc-deprotection. However, this synthetic strategy is quite tedious. (Solid DEL-2)

Fig. 2: Synthetic strategy used for "self-purifying" release of DEL (Solid DEL-2).

Brian M. Paegel has pioneered an alternative solid-phase DNA-encoded library (DEL) synthesis strategy that integrates the "one-bead-one-compound" (OBOC) approach3. In this method, library members are constructed on solid-phase microbeads and linked to DNA tags via a photocleavable linker. This design facilitates the physical isolation and light-triggered release of desired peptide compounds, thereby enabling a broader spectrum of screening modalities, including activity-based assays and cellular phenotypic screening. However, this approach is inherently limited to a library size of 104–106 members, as it relies on the individual screening of compounds on discrete physical beads (Fig. 3). (Solid DEL-3)

Fig. 3: Synthetic strategy used for one-bead-one-compound of DEL (Solid DEL-3).

Collectively, these three solid-phase DEL design paradigms offer new insights into the future of peptide DEL library construction. They suggest that we can strategically leverage emerging technologies to fundamentally enhance peptide DEL library quality. Aligning with this evolving paradigm, HitGen is also exploring the introduction of novel solid-phase methodologies to elevate the quality of our peptide libraries. We anticipate that, in the near future, these innovations will be successfully translated into practice, pointing a new direction for the next generation of DEL synthesis.

Reference:

1. He Q, Wang Y, Tang X, et al. Enhanced screening via a pure DNA-encoded peptide library enabled by an Fmoc modification. Proc Natl Acad Sci U S A. 2026;123(8):e2524999123. doi:10.1073/pnas.2524999123

2. Keller M, Petrov D, Gloger A, et al. Highly pure DNA-encoded chemical libraries by dual-linker solid-phase synthesis. Science. 2024;384(6701):1259-1265. doi:10.1126/science.adn3412

3. Dixit A, Paegel BM. Solid-phase DNA-encoded library synthesis: a master builder's instructions. Nat Protoc. 2026;21(2):542-581. doi:10.1038/s41596-025-01190-4