Chemistry Informer Libraries: Conception, Early Experience, and Role in the Future of Cheminformatics

Cheminformatic Analysis of Core-Atom Transformations in Pharmaceutically Relevant Heteroaromatics

Heteroaromatics are the basis for many pharmaceuticals. The ability to modify these structures through selective core-atom transformations, or "skeletal edits", can dramatically expand the landscape for drug discovery and development. However, despite the importance of core-atom modifications, the quantitative impact of such transformations on accessible chemical space remains undefined. Here, we report a cheminformatic platform to analyze which skeletal edits would most increase access to novel chemical space. This study underscores the significance of emerging single and multiple core-atom transformations of heteroaromatics in enhancing chemical diversity, for example, at a late-stage of a drug discovery campaign. Our findings provide a quantitative framework for prioritizing core-atom modifications in heteroaromatic structural motifs, calling for the development of new methods to achieve these types of transformations.

Recent Green and Sustainable Pd-Catalyzed Aminations

Pd-catalyzed aminations, a powerful and commonly employed method of CN bond construction, often rely on unsustainable technologies that utilize egregious organic solvents, high temperatures, long reaction times, and high catalyst loadings, especially of palladium. Only recently there has been a shift toward far greener protocols based on recyclable aqueous media as well as nontraditional organic solvents. In addition, alternatives to batch methods for preparing the same amines have appeared, such as continuous flow and mechanochemistry, which also offer safe and sustainable means of chemical synthesis associated with targets in the fine chemicals industry.

Designing Target-specific Data Sets for Regioselectivity Predictions on Complex Substrates

The development of machine learning models to predict the regioselectivity of C(sp3)-H functionalization reactions is reported. A data set for dioxirane oxidations was curated from the literature and used to generate a model to predict the regioselectivity of C-H oxidation. To assess whether smaller, intentionally designed data sets could provide accuracy on complex targets, a series of acquisition functions were developed to select the most informative molecules for the specific target. Active learning-based acquisition functions that leverage predicted reactivity and model uncertainty were found to outperform those based on molecular and site similarity alone. The use of acquisition functions for data set elaboration significantly reduced the number of data points needed to perform accurate prediction, and it was found that smaller, machine-designed data sets can give accurate predictions when larger, randomly selected data sets fail. Finally, the workflow was experimentally validated on five complex substrates and shown to be applicable to predicting the regioselectivity of arene C-H radical borylation. These studies provide a quantitative alternative to the intuitive extrapolation from "model substrates" that is frequently used to estimate reactivity on complex molecules.

Recommending reaction conditions with label ranking

Pinpointing effective reaction conditions can be challenging, even for reactions with significant precedent. Herein, models that rank reaction conditions are introduced as a conceptually new means for prioritizing experiments, distinct from the mainstream approach of yield regression. Specifically, label ranking, which operates using input features only from substrates, will be shown to better generalize to new substrates than prior models. Evaluation on practical reaction condition selection scenarios - choosing from either 4 or 18 conditions and datasets with or without missing reactions - demonstrates label ranking's utility. Ranking aggregation through Borda's method and relative simplicity are key features of label ranking to achieve consistent high performance.

Late-stage installation and functionalization of alkyl pyridiniums: a general HTE amenable strategy to access diverse aryl alanine containing macrocyclic peptides

This manuscript describes a strategy to readily access diverse aryl and homoaryl alanine-containing pharmaceutically relevant macrocyclic peptides. A two-step sequence involving the late-stage installation of the pyridinium functionality on macrocyclic peptides followed by reductive couplings was implemented. These transformations are amenable to microscale high-throughput experimentation (HTE) and enable rapid access to aryl alanine-containing macrocyclic peptides that would otherwise be inaccessible via solid-phase peptide synthesis using commercially available amino acids. Numerous aryl and heteroaryl derivatives can be effectively used in these reactions. In addition, a systematic investigation was undertaken using an "informer" set of macrocyclic peptides which revealed the compatibility of the late-stage diversification with peptides containing diverse side chain functionalities.

Homogenous Palladium-Catalyzed Dehalogenative Deuteration and Tritiation of Aryl Halides with D2/T2 Gas

Hydrogen isotopically labeled compounds have extensive utility across diverse domains, especially in drug discovery and development. However, synthesis of the labeled compounds with exclusive site selectivity and/or high isotope incorporation is challenging. One widely employed method is heterogeneous palladium(0)-catalyzed (such as Pd/C) dehalogenative deuteration and tritiation with D2/T2 gas. While commonly used, the method faces two long-standing challenges related to insufficient isotope incorporation and functional group tolerance, particularly with aryl bromides and chlorides. These long-standing issues pose a substantial obstacle in the synthesis of deuterated drug molecules and high-specific-activity tritium tracers. Herein, we present a novel palladium catalytic system using Zn(OAc)2 as an additive, enabling novel homogenous dehalogenative deuteration/tritiation using D2/T2 gas. Under mild reaction conditions, a wide range of drug-like aryl halides and pseudohalides undergo selective deuteration with complete isotope incorporation. The reaction displays excellent compatibility with diverse functional groups, including multiple bonds and O/N-benzyl, and cyano groups, which are frequently problematic in the Pd/C reactions. Furthermore, this method was successfully applied to the tritiation of four halogenated pharmaceutically relevant molecules, resulting in predictable high specific activity per halogen atom (26.5-27.7 Ci/mmol). Notably, the developed system allows gram-scale preparation of a deuterium-containing intermediate, a crucial step in synthesizing a deuterium-labeled drug molecule. A key intermediate, Pd(Ar)OAc, is proposed to activate hydrogen gas during dehalogenative deuteration and tritiation, and Zn(OAc)2 plays an essential role in inhibiting Pd poisoning by halides.

Synthesis and Suzuki-Miyaura Cross-Coupling of Alkyl Amine-Boranes. A Boryl Radical-Enabled Strategy

Alkyl organoborons are powerful materials for the construction of C(sp3)-C(sp2) bonds, predominantly via Suzuki-Miyaura cross-coupling. These species are generally assembled using 2-electron processes that harness the ability of boron reagents to act as both electrophiles and nucleophiles. Herein, we demonstrate an alternative borylation strategy based on the reactivity of amine-ligated boryl radicals. This process features the use of a carboxylic acid containing amine-ligated borane that acts as boryl radical precursor for photoredox oxidation and decarboxylation. The resulting amine-ligated boryl radical undergoes facile addition to styrenes and imines through radical-polar crossover manifolds. This delivers a new class of sp3-organoborons that are stable solids and do not undergo protodeboronation. These novel materials include unprotected α-amino derivatives that are generally unstable. Crucially, these aliphatic organoboron species can be directly engaged in Suzuki-Miyaura cross-couplings with structurally complex aryl halides. Preliminary studies suggest that they enable slow-release of the corresponding and often difficult to handle alkyl boronic acids.

Standardizing Substrate Selection: A Strategy toward Unbiased Evaluation of Reaction Generality

With over 10,000 new reaction protocols arising every year, only a handful of these procedures transition from academia to application. A major reason for this gap stems from the lack of comprehensive knowledge about a reaction's scope, i.e., to which substrates the protocol can or cannot be applied. Even though chemists invest substantial effort to assess the scope of new protocols, the resulting scope tables involve significant biases, reducing their expressiveness. Herein we report a standardized substrate selection strategy designed to mitigate these biases and evaluate the applicability, as well as the limits, of any chemical reaction. Unsupervised learning is utilized to map the chemical space of industrially relevant molecules. Subsequently, potential substrate candidates are projected onto this universal map, enabling the selection of a structurally diverse set of substrates with optimal relevance and coverage. By testing our methodology on different chemical reactions, we were able to demonstrate its effectiveness in finding general reactivity trends by using a few highly representative examples. The developed methodology empowers chemists to showcase the unbiased applicability of novel methodologies, facilitating their practical applications. We hope that this work will trigger interdisciplinary discussions about biases in synthetic chemistry, leading to improved data quality.

Late-stage meta-C-H alkylation of pharmaceuticals to modulate biological properties and expedite molecular optimisation in a single step

Catalysed C-H activation has emerged as a transformative platform for molecular synthesis and provides new opportunities in drug discovery by late-stage functionalisation (LSF) of complex molecules. Notably, small aliphatic motifs have gained significant interest in medicinal chemistry for their beneficial properties and applications as sp3-rich functional group bioisosteres. In this context, we disclose a versatile strategy with broad applicability for the ruthenium-catalysed late-stage meta-C(sp2)-H alkylation of pharmaceuticals. This general protocol leverages numerous directing groups inherently part of bioactive scaffolds to selectivity install a variety of medicinally relevant bifunctional alkyl units within drug compounds. Our strategy enables the direct modification of unprotected lead structures to quickly generate an array of pharmaceutically useful analogues without resorting to de novo syntheses. Moreover, productive late-stage modulation of key biological characteristics of drug candidates upon remote C-H alkylation proves viable, highlighting the major benefits of our approach to offer in drug development programmes.

Development of High-Throughput Experimentation Approaches for Rapid Radiochemical Exploration

Positron emission tomography is a widely used imaging platform for studying physiological processes. Despite the proliferation of modern synthetic methodologies for radiolabeling, the optimization of these reactions still primarily relies on inefficient one-factor-at-a-time approaches. High-throughput experimentation (HTE) has proven to be a powerful approach for optimizing reactions in many areas of chemical synthesis. However, to date, HTE has rarely been applied to radiochemistry. This is largely because of the short lifetime of common radioisotopes, which presents major challenges for efficient parallel reaction setup and analysis using standard equipment and workflows. Herein, we demonstrate an effective HTE workflow and apply it to the optimization of copper-mediated radiofluorination of pharmaceutically relevant boronate ester substrates. The workflow utilizes commercial equipment and allows for rapid analysis of reactions for optimizing reactions, exploring chemical space using pharmaceutically relevant aryl boronates for radiofluorinations, and constructing large radiochemistry data sets.

Bridging the information gap in organic chemical reactions

The varying quality of scientific reports is a well-recognized problem and often results from a lack of standardization and transparency in scientific publications. This situation ultimately leads to prominent complications such as reproducibility issues and the slow uptake of newly developed synthetic methods for pharmaceutical and agrochemical applications. In recent years, various impactful approaches have been advocated to bridge information gaps and to improve the quality of experimental protocols in synthetic organic publications. Here we provide a critical overview of these strategies and present the reader with a versatile set of tools to augment their standard procedures. We formulate eight principles to improve data management in scientific publications relating to data standardization, reproducibility and evaluation, and encourage scientists to go beyond current publication standards. We are aware that this is a substantial effort, but we are convinced that the resulting improved data situation will greatly benefit the progress of chemistry.

Couple-close construction of polycyclic rings from diradicals

Heteroarenes are ubiquitous motifs in bioactive molecules, conferring favourable physical properties when compared to their arene counterparts1-3. In particular, semisaturated heteroarenes possess attractive solubility properties and a higher fraction of sp3 carbons, which can improve binding affinity and specificity. However, these desirable structures remain rare owing to limitations in current synthetic methods4-6. Indeed, semisaturated heterocycles are laboriously prepared by means of non-modular fit-for-purpose syntheses, which decrease throughput, limit chemical diversity and preclude their inclusion in many hit-to-lead campaigns7-10. Herein, we describe a more intuitive and modular couple-close approach to build semisaturated ring systems from dual radical precursors. This platform merges metallaphotoredox C(sp2)-C(sp3) cross-coupling with intramolecular Minisci-type radical cyclization to fuse abundant heteroaryl halides with simple bifunctional feedstocks, which serve as the diradical synthons, to rapidly assemble a variety of spirocyclic, bridged and substituted saturated ring types that would be extremely difficult to make by conventional methods. The broad availability of the requisite feedstock materials allows sampling of regions of underexplored chemical space. Reagent-controlled radical generation leads to a highly regioselective and stereospecific annulation that can be used for the late-stage functionalization of pharmaceutical scaffolds, replacing lengthy de novo syntheses.

Identifying general reaction conditions by bandit optimization

Reaction conditions that are generally applicable to a wide variety of substrates are highly desired, especially in the pharmaceutical and chemical industries1-6. Although many approaches are available to evaluate the general applicability of developed conditions, a universal approach to efficiently discover these conditions during optimizations is rare. Here we report the design, implementation and application of reinforcement learning bandit optimization models7-10 to identify generally applicable conditions by efficient condition sampling and evaluation of experimental feedback. Performance benchmarking on existing datasets statistically showed high accuracies for identifying general conditions, with up to 31% improvement over baselines that mimic state-of-the-art optimization approaches. A palladium-catalysed imidazole C-H arylation reaction, an aniline amide coupling reaction and a phenol alkylation reaction were investigated experimentally to evaluate use cases and functionalities of the bandit optimization model in practice. In all three cases, the reaction conditions that were most generally applicable yet not well studied for the respective reaction were identified after surveying less than 15% of the expert-designed reaction space.

Enabling late-stage drug diversification by high-throughput experimentation with geometric deep learning

Late-stage functionalization is an economical approach to optimize the properties of drug candidates. However, the chemical complexity of drug molecules often makes late-stage diversification challenging. To address this problem, a late-stage functionalization platform based on geometric deep learning and high-throughput reaction screening was developed. Considering borylation as a critical step in late-stage functionalization, the computational model predicted reaction yields for diverse reaction conditions with a mean absolute error margin of 4-5%, while the reactivity of novel reactions with known and unknown substrates was classified with a balanced accuracy of 92% and 67%, respectively. The regioselectivity of the major products was accurately captured with a classifier F-score of 67%. When applied to 23 diverse commercial drug molecules, the platform successfully identified numerous opportunities for structural diversification. The influence of steric and electronic information on model performance was quantified, and a comprehensive simple user-friendly reaction format was introduced that proved to be a key enabler for seamlessly integrating deep learning and high-throughput experimentation for late-stage functionalization.

A General Strategy for N-(Hetero)arylpiperidine Synthesis Using Zincke Imine Intermediates

Methods to synthesize diverse collections of substituted piperidines are valuable due to the prevalence of this heterocycle in pharmaceutical compounds. Here, we present a general strategy to access N-(hetero)arylpiperidines using a pyridine ring-opening and ring-closing approach via Zincke imine intermediates. This process generates pyridinium salts from a wide variety of substituted pyridines and (heteroaryl)anilines; hydrogenation reactions and nucleophilic additions then access the N-(hetero)arylpiperidine derivatives. We successfully applied high-throughput experimentation (HTE) using pharmaceutically relevant pyridines and (heteroaryl)anilines as inputs and developed a one-pot process using anilines as nucleophiles in the pyridinium salt-forming processes. This strategy is viable for generating piperidine libraries and applications such as the convergent coupling of complex fragments.

When Yield Prediction Does Not Yield Prediction: An Overview of the Current Challenges

Machine Learning (ML) techniques face significant challenges when predicting advanced chemical properties, such as yield, feasibility of chemical synthesis, and optimal reaction conditions. These challenges stem from the high-dimensional nature of the prediction task and the myriad essential variables involved, ranging from reactants and reagents to catalysts, temperature, and purification processes. Successfully developing a reliable predictive model not only holds the potential for optimizing high-throughput experiments but can also elevate existing retrosynthetic predictive approaches and bolster a plethora of applications within the field. In this review, we systematically evaluate the efficacy of current ML methodologies in chemoinformatics, shedding light on their milestones and inherent limitations. Additionally, a detailed examination of a representative case study provides insights into the prevailing issues related to data availability and transferability in the discipline.

Dataset Design for Building Models of Chemical Reactivity

Models can codify our understanding of chemical reactivity and serve a useful purpose in the development of new synthetic processes via, for example, evaluating hypothetical reaction conditions or in silico substrate tolerance. Perhaps the most determining factor is the composition of the training data and whether it is sufficient to train a model that can make accurate predictions over the full domain of interest. Here, we discuss the design of reaction datasets in ways that are conducive to data-driven modeling, emphasizing the idea that training set diversity and model generalizability rely on the choice of molecular or reaction representation. We additionally discuss the experimental constraints associated with generating common types of chemistry datasets and how these considerations should influence dataset design and model building.

Branched-Selective Cross-Electrophile Coupling of 2-Alkyl Aziridines and (Hetero)aryl Iodides Using Ti/Ni Catalysis

The arylation of 2-alkyl aziridines by nucleophilic ring-opening or transition-metal-catalyzed cross-coupling enables facile access to biologically relevant β-phenethylamine derivatives. However, both approaches largely favor C-C bond formation at the less-substituted carbon of the aziridine, thus enabling access to only linear products. Consequently, despite the attractive bond disconnection that it poses, the synthesis of branched arylated products from 2-alkyl aziridines has remained inaccessible. Herein, we address this long-standing challenge and report the first branched-selective cross-coupling of 2-alkyl aziridines with aryl iodides. This unique selectivity is enabled by a Ti/Ni dual-catalytic system. We demonstrate the robustness of the method by a twofold approach: an additive screening campaign to probe functional group tolerance and a feature-driven substrate scope to study the effect of the local steric and electronic profile of each coupling partner on reactivity. Furthermore, the diversity of this feature-driven substrate scope enabled the generation of predictive reactivity models that guided mechanistic understanding. Mechanistic studies demonstrated that the branched selectivity arises from a TiIII-induced radical ring-opening of the aziridine.

Nickel-Catalyzed Electroreductive Coupling of Alkylpyridinium Salts and Aryl Halides

An electrochemical, nickel-catalyzed reductive coupling of alkylpyridinium salts and aryl halides is reported. High-throughput experimentation (HTE) was employed for rapid reaction optimization and evaluation of a broad scope of pharmaceutically relevant structurally diverse aryl halides, including complex drug-like substrates. In addition, the transformation is compatible with both primary and secondary alkylpyridinium salts with distinct conditions. Mechanistic insights were critical to enhance the efficiency of coupling using secondary alkylpyridinium salts. Systematic comparisons of the electrochemical and non-electrochemical methods revealed the complementary scope and efficiency of the two approaches.

Global reactivity models are impactful in industrial synthesis applications

Artificial Intelligence is revolutionizing many aspects of the pharmaceutical industry. Deep learning models are now routinely applied to guide drug discovery projects leading to faster and improved findings, but there are still many tasks with enormous unrealized potential. One such task is the reaction yield prediction. Every year more than one fifth of all synthesis attempts result in product yields which are either zero or too low. This equates to chemical and human resources being spent on activities which ultimately do not progress the programs, leading to a triple loss when accounting for the cost of opportunity in time wasted. In this work we pre-train a BERT model on more than 16 million reactions from 4 different data sources, and fine tune it to achieve an uncertainty calibrated global yield prediction model. This model is an improvement upon state of the art not just from the increase in pre-train data but also by introducing a new embedding layer which solves a few limitations of SMILES and enables integration of additional information such as equivalents and molecule role into the reaction encoding, the model is called BERT Enriched Embedding (BEE). The model is benchmarked on an open-source dataset against a state-of-the-art synthesis focused BERT showing a near 20-point improvement in r2 score. The model is fine-tuned and tested on an internal company data benchmark, and a prospective study shows that the application of the model can reduce the total number of negative reactions (yield under 5%) ran in Janssen by at least 34%. Lastly, we corroborate the previous results through experimental validation, by directly deploying the model in an on-going drug discovery project and showing that it can also be used successfully as a reagent recommender due to its fast inference speed and reliable confidence estimation, a critical feature for industry application.

Accelerating reaction generality and mechanistic insight through additive mapping

Reaction generality is crucial in determining the overall impact and usefulness of synthetic methods. Typical generalization protocols require a priori mechanistic understanding and suffer when applied to complex, less understood systems. We developed an additive mapping approach that rapidly expands the utility of synthetic methods while generating concurrent mechanistic insight. Validation of this approach on the metallaphotoredox decarboxylative arylation resulted in the discovery of a phthalimide ligand additive that overcomes many lingering limitations of this reaction and has important mechanistic implications for nickel-catalyzed cross-couplings.

Merging Directed C-H Activations with High-Throughput Experimentation: Development of Iridium-Catalyzed C-H Aminations Applicable to Late-Stage Functionalization

Herein, we report an iridium-catalyzed directed C-H amination methodology developed using a high-throughput experimentation (HTE)-based strategy, applicable for the needs of automated modern drug discovery. The informer library approach for investigating the accessible directing group chemical space, in combination with functional group tolerance screening and substrate scope investigations, allowed for the generation of reaction application guidelines to aid future users. Applicability to late-stage functionalization of complex drugs and natural products, in combination with multiple deprotection protocols leading to the desirable aniline matched pairs, serve to demonstrate the utility of the method for drug discovery. Finally, reaction miniaturization to a nanomolar range highlights the opportunities for more sustainable screening with decreased material consumption.

Using Data Science To Guide Aryl Bromide Substrate Scope Analysis in a Ni/Photoredox-Catalyzed Cross-Coupling with Acetals as Alcohol-Derived Radical Sources

Ni/photoredox catalysis has emerged as a powerful platform for C(sp2)-C(sp3) bond formation. While many of these methods typically employ aryl bromides as the C(sp2) coupling partner, a variety of aliphatic radical sources have been investigated. In principle, these reactions enable access to the same product scaffolds, but it can be hard to discern which method to employ because nonstandardized sets of aryl bromides are used in scope evaluation. Herein, we report a Ni/photoredox-catalyzed (deutero)methylation and alkylation of aryl halides where benzaldehyde di(alkyl) acetals serve as alcohol-derived radical sources. Reaction development, mechanistic studies, and late-stage derivatization of a biologically relevant aryl chloride, fenofibrate, are presented. Then, we describe the integration of data science techniques, including DFT featurization, dimensionality reduction, and hierarchical clustering, to delineate a diverse and succinct collection of aryl bromides that is representative of the chemical space of the substrate class. By superimposing scope examples from published Ni/photoredox methods on this same chemical space, we identify areas of sparse coverage and high versus low average yields, enabling comparisons between prior art and this new method. Additionally, we demonstrate that the systematically selected scope of aryl bromides can be used to quantify population-wide reactivity trends and reveal sources of possible functional group incompatibility with supervised machine learning.

Enantioselective sensing of insect pheromones in water

An arrayed combination of water-soluble deep cavitands and cationic dyes has been shown to optically sense insect pheromones at micromolar concentration in water. Machine learning approaches were used to optimize the most effective array components, which allows differentiation between small structural differences in targets, including between different diastereomers, even though the pheromones have no innate chromophore. When combined with chiral additives, enantiodiscrimination is possible, dependent on the size and shape of the pheromone.

Lactobacillus plantarum alleviates irradiation-induced intestinal injury by activation of FXR-FGF15 signaling in intestinal epithelia

Abdominal irradiation (IR) may destroy the intestinal mucosal barrier, leading to severe intestinal infection and multiple organ dysfunction syndromes. The role of intestinal microbiota in the development of IR-induced intestinal injury remains largely unknown. Herein, we reported that abdominal IR altered the composition of the microbiota and reduced the abundance and diversity of the gut microbiome. Alterations of bacteria, in particular reduction of Lactobacillus, played a critical role in IR-induced intestinal injury. Fecal microbiota transplant (FMT) from normal mice or administration of Lactobacillus plantarum to intestinal microbiota-eliminated mice substantially reduced IR-induced intestinal damage and prevented mice from IR-induced death. We further characterized that L. plantarum activated the farnesoid X receptor (FXR) - fibroblast growth factor 15 (FGF15) signaling in intestinal epithelial cells and hence promoted DNA-damage repair. Application of GW4064, an activator of FXR, to microbiota eliminated mice markedly mitigated IR-induced intestinal damage, reduced intestinal epithelial cell death and promoted the survival of IR mice. In contrast, suppression of FXR with Gly-β-MCA, a bile acid and an intestine-selective and high-affinity FXR inhibitor, abrogated L. Plantarum-mediated protection on the ileum of IR mice. Taken together, our findings not only provide new insights into the role of intestinal flora in radiation-induced intestinal injury but also shed new light on the application of probiotics for the protection of radiation-damaged individuals.

The Open Reaction Database

Chemical reaction data in journal articles, patents, and even electronic laboratory notebooks are currently stored in various formats, often unstructured, which presents a significant barrier to downstream applications, including the training of machine-learning models. We present the Open Reaction Database (ORD), an open-access schema and infrastructure for structuring and sharing organic reaction data, including a centralized data repository. The ORD schema supports conventional and emerging technologies, from benchtop reactions to automated high-throughput experiments and flow chemistry. The data, schema, supporting code, and web-based user interfaces are all publicly available on GitHub. Our vision is that a consistent data representation and infrastructure to support data sharing will enable downstream applications that will greatly improve the state of the art with respect to computer-aided synthesis planning, reaction prediction, and other predictive chemistry tasks.

Machine Learning Aids Classification and Discrimination of Noncanonical DNA Folding Motifs by an Arrayed Host:Guest Sensing System

An arrayed host:guest fluorescence sensor system can discriminate among and classify multiple different noncanonical DNA structures by exploiting selective molecular recognition. The sensor is highly selective and can discriminate between folds as similar as native G-quadruplexes and those with bulges or vacancies. The host and guest can form heteroternary complexes with DNA strands, with the host acting as mediator between the DNA and dye, modulating the emission. By applying machine learning algorithms to the sensing data, prediction of the folding state of unknown DNA strands is possible with high fidelity.