Computational advances in combating colloidal aggregation in drug discovery

Boosting the Accuracy and Chemical Space Coverage of the Detection of Small Colloidal Aggregating Molecules Using the BAD Molecule Filter

The ability to conduct effective high throughput screening (HTS) campaigns in drug discovery is often hampered by the detection of false positives in these assays due to small colloidally aggregating molecules (SCAMs). SCAMs can produce artifactual hits in HTS by nonspecific inhibition of the protein target. In this work, we present a new computational prediction tool for detecting SCAMs based on their 2D chemical structure. The tool, called the boosted aggregation detection (BAD) molecule filter, employs decision tree ensemble methods, namely, the CatBoost classifier and the light gradient-boosting machine, to significantly improve the detection of SCAMs. In developing the filter, we explore models trained on individual data sets, a consensus approach using these models, and, third, a merged data set approach, each tailored for specific drug discovery needs. The individual data set method emerged as most effective, achieving 93% sensitivity and 90% specificity, outperforming existing state-of-the-art models by 20 and 5%, respectively. The consensus models offer broader chemical space coverage, exceeding 90% for all testing sets. This feature is an important aspect particularly for early stage medicinal chemistry projects, and provides information on applicability domain. Meanwhile, the merged data set models demonstrated robust performance, with a notable sensitivity of 79% in the comprehensive 10-fold cross-validation test set. A SHAP analysis of model features indicates the importance of hydrophobicity and molecular complexity as primary factors influencing the aggregation propensity. The BAD molecule filter is readily accessible for the public usage on https://molmodlab-aau.com/Tools.html. This filter provides a new, more robust tool for aggregate prediction in the early stages of drug discovery to optimize hit rates and reduce associated testing and validation overheads.

Predictions of Colloidal Molecular Aggregation Using AI/ML Models

To facilitate the triage of hits from small molecule screens, we have used various AI/ML techniques and experimentally observed data sets to build models aimed at predicting colloidal aggregation of small organic molecules in aqueous solution. We have found that Naïve Bayesian and deep neural networks outperform logistic regression, recursive partitioning tree, support vector machine, and random forest techniques by having the lowest balanced error rate (BER) for the test set. Derived predictive classification models consistently and successfully discriminated aggregator molecules from nonaggregator hits. An analysis of molecular descriptors in favor of colloidal aggregation confirms previous observations (hydrophobicity, molecular weight, and solubility) in addition to undescribed molecular descriptors such as the fraction of sp3 carbon atoms (Fsp3), and electrotopological state of hydroxyl groups (ES_Sum_sOH). Naïve Bayesian modeling and scaffold tree analysis have revealed chemical features/scaffolds contributing the most to colloidal aggregation and nonaggregation, respectively. These results highlight the importance of scaffolds with high Fsp3 values in promoting nonaggregation. Matched molecular pair analysis (MMPA) has also deciphered context-dependent substitutions, which can be used to design nonaggregator molecules. We found that most matched molecular pairs have a neutral effect on aggregation propensity. We have prospectively applied our predictive models to assist in chemical library triage for optimal plate selection diversity and purchase for high throughput screening (HTS) in drug discovery projects.

Tackling assay interference associated with small molecules

Biochemical and cell-based assays are essential to discovering and optimizing efficacious and safe drugs, agrochemicals and cosmetics. However, false assay readouts stemming from colloidal aggregation, chemical reactivity, chelation, light signal attenuation and emission, membrane disruption, and other interference mechanisms remain a considerable challenge in screening synthetic compounds and natural products. To address assay interference, a range of powerful experimental approaches are available and in silico methods are now gaining traction. This Review begins with an overview of the scope and limitations of experimental approaches for tackling assay interference. It then focuses on theoretical methods, discusses strategies for their integration with experimental approaches, and provides recommendations for best practices. The Review closes with a summary of the critical facts and an outlook on potential future developments.

Fragment-based drug nanoaggregation reveals drivers of self-assembly

Drug nanoaggregates are particles that can deleteriously cause false positive results during drug screening efforts, but alternatively, they may be used to improve pharmacokinetics when developed for drug delivery purposes. The structural features of molecules that drive nanoaggregate formation remain elusive, however, and the prediction of intracellular aggregation and rational design of nanoaggregate-based carriers are still challenging. We investigate nanoaggregate self-assembly mechanisms using small molecule fragments to identify the critical molecular forces that contribute to self-assembly. We find that aromatic groups and hydrogen bond acceptors/donors are essential for nanoaggregate formation, suggesting that both π-π stacking and hydrogen bonding are drivers of nanoaggregation. We apply structure-assembly-relationship analysis to the drug sorafenib and discover that nanoaggregate formation can be predicted entirely using drug fragment substructures. We also find that drug nanoaggregates are stabilized in an amorphous core-shell structure. These findings demonstrate that rational design can address intracellular aggregation and pharmacologic/delivery challenges in conventional and fragment-based drug development processes.

Rigidity-Induced Controlled Aggregation of Binary Colloids

Here, we report the proof-of-concept for controlled aggregation in a binary colloidal system. The binary systems are studied by varying bond flexibility of only one species, while the other species' bonds remain fully flexible. By establishing the underlying relation between gelation and bond rigidity, we demonstrate how the interplay among bond flexibility, critical concentration, and packing volume fraction influenced the aggregation kinetics. Our result shows that rigidity in bonds increases the critical concentration for gels to be formed in the binary mixture. Furthermore, the average number of bonded neighbor analyses reveal the influence of bond rigidity both above and below critical concentrations and show that variation in bond flexibility in only one species alters the kinetics of aggregation of both species. This finding improves our understanding of colloidal aggregation in soft and biological systems.

Lies and Liabilities: Computational Assessment of High-Throughput Screening Hits to Identify Artifact Compounds

Hits from high-throughput screening (HTS) of chemical libraries are often false positives due to their interference with assay detection technology. In response, we generated the largest publicly available library of chemical liabilities and developed "Liability Predictor," a free web tool to predict HTS artifacts. More specifically, we generated, curated, and integrated HTS data sets for thiol reactivity, redox activity, and luciferase (firefly and nano) activity and developed and validated quantitative structure-interference relationship (QSIR) models to predict these nuisance behaviors. The resulting models showed 58-78% external balanced accuracy for 256 external compounds per assay. QSIR models developed and validated herein identify nuisance compounds among experimental hits more reliably than do popular PAINS filters. Both the models and the curated data sets were implemented in "Liability Predictor," publicly available at https://liability.mml.unc.edu/. "Liability Predictor" may be used as part of chemical library design or for triaging HTS hits.

Prediction of cancer nanomedicines self-assembled from meta-synergistic drug pairs

Combination therapy is widely used in cancer medicine due to the benefits of drug synergy and the reduction of acquired resistance. To minimize emergent toxicities, nanomedicines containing drug combinations are being developed, and they have shown encouraging results. However, developing multi-drug loaded nanoparticles is highly complex and lacks predictability. Previously, it was shown that single drugs can self-assemble with near-infrared dye, IR783, to form cancer-targeted nanoparticles. A structure-based predictive model showed that only 4% of the drug space self-assembles with IR783. Here, we mapped the self-assembly outcomes of 77 small molecule drugs and drug pairs with IR783. We found that the small molecule drug space can be divided into five types, and type-1 drugs self-assemble with three out of four possible drug types that do not form stable nanoparticles. To predict the self-assembly outcome of any drug pair, we developed a machine learning model based on decision trees, which was trained and tested with F1-scores of 89.3% and 87.2%, respectively. We used literature text mining to capture drug pairs with biological synergy together with synergistic chemical self-assembly and generated a database with 1985 drug pairs for 70 cancers. We developed an online search tool to identify cancer-specific, meta-synergistic drug pairs (both chemical and biological synergism) and validated three different pairs in vitro. Lastly, we discovered a novel meta-synergistic pair, bortezomib-cabozantinib, which formed stable nanoparticles with improved biodistribution, efficacy, and reduced toxicity, even over single drugs, in an in vivo model of head and neck cancer.

Using Machine Learning to make nanomaterials sustainable

Sustainable development is a key challenge for contemporary human societies; failure to achieve sustainability could threaten human survival. In this review article, we illustrate how Machine Learning (ML) could support more sustainable development, covering the basics of data gathering through each step of the Environmental Risk Assessment (ERA). The literature provides several examples showing how ML can be employed in most steps of a typical ERA.A key observation is that there are currently no clear guidance for using such autonomous technologies in ERAs or which standards/checks are required. Steering thus seems to be the most important task for supporting the use of ML in the ERA of nano- and smart-materials. Resources should be devoted to developing a strategy for implementing ML in ERA with a strong emphasis on data foundations, methodologies, and the related sensitivities/uncertainties. We should recognise historical errors and biases (e.g., in data) to avoid embedding them during ML programming.

Predicting drug-target binding affinity through molecule representation block based on multi-head attention and skip connection

Exiting computational models for drug-target binding affinity prediction have much room for improvement in prediction accuracy, robustness and generalization ability. Most deep learning models lack interpretability analysis and few studies provide application examples. Based on these observations, we presented a novel model named Molecule Representation Block-based Drug-Target binding Affinity prediction (MRBDTA). MRBDTA is composed of embedding and positional encoding, molecule representation block and interaction learning module. The advantages of MRBDTA are reflected in three aspects: (i) developing Trans block to extract molecule features through improving the encoder of transformer, (ii) introducing skip connection at encoder level in Trans block and (iii) enhancing the ability to capture interaction sites between proteins and drugs. The test results on two benchmark datasets manifest that MRBDTA achieves the best performance compared with 11 state-of-the-art models. Besides, through replacing Trans block with single Trans encoder and removing skip connection in Trans block, we verified that Trans block and skip connection could effectively improve the prediction accuracy and reliability of MRBDTA. Then, relying on multi-head attention mechanism, we performed interpretability analysis to illustrate that MRBDTA can correctly capture part of interaction sites between proteins and drugs. In case studies, we firstly employed MRBDTA to predict binding affinities between Food and Drug Administration-approved drugs and severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) replication-related proteins. Secondly, we compared true binding affinities between 3C-like proteinase and 185 drugs with those predicted by MRBDTA. The final results of case studies reveal reliable performance of MRBDTA in drug design for SARS-CoV-2.

Automated discovery of nanomaterials via drug aggregation induced emission

Nanoformulations of small molecule drugs are essential to effectively deliver them and treat a wide range of diseases. They are normally complex to develop, lack predictability, and exhibit low drug loading. Recently, nanoparticles made via co-assembly of hydrophobic drugs and organic dyes, exhibited drug-loading of up to 90% with high predictability from the drug structure. However, these particles have relatively short stability and can formulate only a small fraction of the drug space. Here, we developed an automated workflow to synthesize and select novel dye stabilizers, based on their ability to inhibit drug aggregation-induced emission (AIE). We first screened and identified 10 drugs with previously unknown strong AIE activity and exploited this trait to automatically synthesize and select a new ultra-stabilizer named R595. Interestingly, it shares several synthetic similarities and advantages with polydopamine. We found that R595 is superior to myriad types of excipients and solubilizers such as cyclodextrins, poloxamers, albumin, and previously published organic dyes, in both long-term stability and drug compatibility. We investigated the biodistribution, pharmacokinetics, safety and efficacy of the AIEgenic MEK inhibitor trametinib-R595 nanoparticles in vitro and in vivo and demonstrated that they are non-toxic and effective in KRAS driven colon and lung cancer models.

Evaluation guidelines for machine learning tools in the chemical sciences

Machine learning (ML) promises to tackle the grand challenges in chemistry and speed up the generation, improvement and/or ordering of research hypotheses. Despite the overarching applicability of ML workflows, one usually finds diverse evaluation study designs. The current heterogeneity in evaluation techniques and metrics leads to difficulty in (or the impossibility of) comparing and assessing the relevance of new algorithms. Ultimately, this may delay the digitalization of chemistry at scale and confuse method developers, experimentalists, reviewers and journal editors. In this Perspective, we critically discuss a set of method development and evaluation guidelines for different types of ML-based publications, emphasizing supervised learning. We provide a diverse collection of examples from various authors and disciplines in chemistry. While taking into account varying accessibility across research groups, our recommendations focus on reporting completeness and standardizing comparisons between tools. We aim to further contribute to improved ML transparency and credibility by suggesting a checklist of retro-/prospective tests and dissecting their importance. We envisage that the wide adoption and continuous update of best practices will encourage an informed use of ML on real-world problems related to the chemical sciences.

Isometric Stratified Ensembles: A Partial and Incremental Adaptive Applicability Domain and Consensus-Based Classification Strategy for Highly Imbalanced Data Sets with Application to Colloidal Aggregation

Partial and incremental stratification analysis of a quantitative structure-interference relationship (QSIR) is a novel strategy intended to categorize classification provided by machine learning techniques. It is based on a 2D mapping of classification statistics onto two categorical axes: the degree of consensus and level of applicability domain. An internal cross-validation set allows to determine the statistical performance of the ensemble at every 2D map stratum and hence to define isometric local performance regions with the aim of better hit ranking and selection. During training, isometric stratified ensembles (ISE) applies a recursive decorrelated variable selection and considers the cardinal ratio of classes to balance training sets and thus avoid bias due to possible class imbalance. To exemplify the interest of this strategy, three different highly imbalanced PubChem pairs of AmpC β-lactamase and cruzain inhibition assay campaigns of colloidal aggregators and complementary aggregators data set available at the AGGREGATOR ADVISOR predictor web page were employed. Statistics obtained using this new strategy show outperforming results compared to former published tools, with and without a classical applicability domain. ISE performance on classifying colloidal aggregators shows from a global AUC of 0.82, when the whole test data set is considered, up to a maximum AUC of 0.88, when its highest confidence isometric stratum is retained.

Molecular characteristics of several drugs evaluated from solvent/water partition measurements: Solvation parameters and intramolecular hydrogen bond indicator

A wide set of well-known drugs, most of them included in the Abraham´s reference database, covering a wide variety of chemical structures and therapeutical functionalities were chosen in order to determine some molecular properties from solvent/water partition measurements. Partition data from aqueous solutions and four different solvents (n-dodecane, toluene, chloroform and n-octanol) were measured and reported. From them, Abraham´s molecular descriptors of selected compounds (A, B and S, accounting for hydrogen bond donor, hydrogen bond acceptor and dipolarity/polaritzability, respectively) were estimated. A and B values derived from the experimental measurements strongly agree with the tabulated ones showing the suitability of the used procedure to achieve reliable values for new molecules. However, obtained S values differ from those previously reported for several compounds. Moreover, values for a new indicator of the propensity to form intramolecular hydrogen bonds (Δlog P_oct-tol) were estimated from the experimental data and also calculated according to both, the Abraham´s model and the molecular structures (SMD). The quality of both series of calculated descriptors was evaluated by contrast with the experimental values and satisfactory results were obtained in both instances. Thus, the Abraham´s way is useful when molecular descriptors are available but very good estimations can be achieved by SMD, which only requires the drug´s molecular structure.

Discovery of Highly Potent Fusion Inhibitors with Potential Pan-Coronavirus Activity That Effectively Inhibit Major COVID-19 Variants of Concern (VOCs) in Pseudovirus-Based Assays

We report the discovery of several highly potent small molecules with low-nM potency against severe acute respiratory syndrome coronavirus (SARS-CoV; lowest half-maximal inhibitory concentration (IC₅₀: 13 nM), SARS-CoV-2 (IC₅₀: 23 nM), and Middle East respiratory syndrome coronavirus (MERS-CoV; IC₅₀: 76 nM) in pseudovirus-based assays with excellent selectivity index (SI) values (>5000), demonstrating potential pan-coronavirus inhibitory activities. Some compounds showed 100% inhibition against the cytopathic effects (CPE; IC₁₀₀) of an authentic SARS-CoV-2 (US_WA-1/2020) variant at 1.25 µM. The most active inhibitors also potently inhibited variants of concern (VOCs), including the UK (B.1.1.7) and South African (B.1.351) variants and the Delta variant (B.1.617.2) originally identified in India in pseudovirus-based assay. Surface plasmon resonance (SPR) analysis with one potent inhibitor confirmed that it binds to the prefusion SARS-CoV-2 spike protein trimer. These small-molecule inhibitors prevented virus-mediated cell-cell fusion. The absorption, distribution, metabolism, and excretion (ADME) data for one of the most active inhibitors, NBCoV1, demonstrated drug-like properties. An in vivo pharmacokinetics (PK) study of NBCoV1 in rats demonstrated an excellent half-life (t_1/2) of 11.3 h, a mean resident time (MRT) of 14.2 h, and oral bioavailability. We expect these lead inhibitors to facilitate the further development of preclinical and clinical candidates.

Updated Prediction of Aggregators and Assay-Interfering Substructures in Food Compounds

Positive outcomes in biochemical and biological assays of food compounds may appear due to the well-described capacity of some compounds to form colloidal aggregates that adsorb proteins, resulting in their denaturation and loss of function. This phenomenon can lead to wrongly ascribing mechanisms of biological action for these compounds (false positives) as the effect is nonspecific and promiscuous. Similar false positives can show up due to chemical (photo)reactivity, redox cycling, metal chelation, interferences with the assay technology, membrane disruption, etc., which are more frequently observed when the tested molecule has some definite interfering substructures. Although discarding false positives can be achieved experimentally, it would be very useful to have in advance a prognostic value for possible aggregation and/or interference based only in the chemical structure of the compound tested in order to be aware of possible issues, help in prioritization of compounds to test, design of appropriate assays, etc. Previously, we applied cheminformatic tools derived from the drug discovery field to identify putative aggregators and interfering substructures in a database of food compounds, the FooDB, comprising 26,457 molecules at that time. Here, we provide an updated account of that analysis based on a current, much-expanded version of the FooDB, comprising a total of 70,855 compounds. In addition, we also apply a novel machine learning model (SCAM Detective) to predict aggregators with 46-53% increased accuracies over previous models. In this way, we expect to provide the researchers in the mode of action of food compounds with a much improved, robust, and widened set of putative aggregators and interfering substructures of food compounds.

Discovery of Zafirlukast as a novel SARS-CoV-2 helicase inhibitor using in silico modelling and a FRET-based assay

The coronavirus helicase is an essential enzyme required for viral replication/transcription pathways. Structural studies revealed a sulphate moiety that interacts with key residues within the nucleotide-binding site of the helicase. Compounds with a sulphoxide or a sulphone moiety could interfere with these interactions and consequently inhibit the enzyme. The molecular operating environment (MOE) was used to dock 189 sulphoxide and sulphone-containing FDA-approved compounds to the nucleotide-binding site. Zafirlukast, a leukotriene receptor antagonist used to treat chronic asthma, achieved the lowest docking score at -8.75 kcals/mol. The inhibitory effect of the compounds on the SARS-CoV-2 helicase dsDNA unwinding activity was tested by a FRET-based assay. Zafirlukast was the only compound to inhibit the enzyme (IC50 = 16.3 µM). The treatment of Vero E6 cells with 25 µM zafirlukast prior to SARS-CoV-2 infection decreased the cytopathic effects of SARS-CoV-2 significantly. These results suggest that zafirlukast alleviates SARS-CoV-2 pathogenicity by inhibiting the viral helicase and impairing the viral replication/transcription pathway. Zafirlukast could be clinically developed as a new antiviral treatment for SARS-CoV-2 and other coronavirus diseases. This discovery is based on molecular modelling, in vitro inhibition of the SARS-CoV helicase activity and cell-based SARS-CoV-2 viral replication.

Gains from no real PAINS: Where 'Fair Trial Strategy' stands in the development of multi-target ligands

Compounds that selectively modulate multiple targets can provide clinical benefits and are an alternative to traditional highly selective agents for unique targets. High-throughput screening (HTS) for multitarget-directed ligands (MTDLs) using approved drugs, and fragment-based drug design has become a regular strategy to achieve an ideal multitarget combination. However, the unexpected presence of pan-assay interference compounds (PAINS) suspects in the development of MTDLs frequently results in nonspecific interactions or other undesirable effects leading to artefacts or false-positive data of biological assays. Publicly available filters can help to identify PAINS suspects; however, these filters cannot comprehensively conclude whether these suspects are "bad" or innocent. Additionally, these in silico approaches may inappropriately label a ligand as PAINS. More than 80% of the initial hits can be identified as PAINS by the filters if appropriate biochemical tests are not used resulting in false positive data that are unacceptable for medicinal chemists in manuscript peer review and future studies. Therefore, extensive offline experiments should be used after online filtering to discriminate "bad" PAINS and avoid incorrect evaluation of good scaffolds. We suggest that the use of "Fair Trial Strategy" to identify interesting molecules in PAINS suspects to provide certain structure‒function insight in MTDL development.

Therapeutics-how to treat phase separation-associated diseases

Liquid-liquid phase separation has drawn attention as many neurodegeneration or cancer-associated proteins are able to form liquid membraneless compartments (condensates) by liquid-liquid phase separation. Furthermore, there is rapidly growing evidence that disease-associated mutation or post-translational modification of these proteins causes aberrant location, composition or physical properties of the condensates. It is ambiguous whether aberrant condensates are always causative in disease mechanisms, however they are likely promising potential targets for therapeutics. The conceptual framework of liquid-liquid phase separation provides opportunities for novel therapeutic approaches. This review summarises how the extensive recent advances in understanding control of nucleation, growth and composition of condensates by protein post-translational modification has revealed many possibilities for intervention by conventional small molecule enzyme inhibitors. This includes the first proof-of-concept examples. However, understanding membraneless organelle formation as a physical chemistry process also highlights possible physicochemical mechanisms of intervention. There is huge demand for innovation in drug development, especially for challenging diseases of old age including neurodegeneration and cancer. The conceptual framework of liquid-liquid phase separation provides a new paradigm for thinking about modulating protein function and is very different from enzyme lock-and-key or structured binding site concepts and presents new opportunities for innovation.

Identifying SARS-CoV-2 antiviral compounds by screening for small molecule inhibitors of nsp13 helicase

The coronavirus disease 2019 (COVID-19) pandemic, which is caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), is a global public health challenge. While the efficacy of vaccines against emerging and future virus variants remains unclear, there is a need for therapeutics. Repurposing existing drugs represents a promising and potentially rapid opportunity to find novel antivirals against SARS-CoV-2. The virus encodes at least nine enzymatic activities that are potential drug targets. Here, we have expressed, purified and developed enzymatic assays for SARS-CoV-2 nsp13 helicase, a viral replication protein that is essential for the coronavirus life cycle. We screened a custom chemical library of over 5000 previously characterized pharmaceuticals for nsp13 inhibitors using a fluorescence resonance energy transfer-based high-throughput screening approach. From this, we have identified FPA-124 and several suramin-related compounds as novel inhibitors of nsp13 helicase activity in vitro. We describe the efficacy of these drugs using assays we developed to monitor SARS-CoV-2 growth in Vero E6 cells.

Computationally guided high-throughput design of self-assembling drug nanoparticles

Nanoformulations of therapeutic drugs are transforming our ability to effectively deliver and treat a myriad of conditions. Often, however, they are complex to produce and exhibit low drug loading, except for nanoparticles formed via co-assembly of drugs and small molecular dyes, which display drug-loading capacities of up to 95%. There is currently no understanding of which of the millions of small-molecule combinations can result in the formation of these nanoparticles. Here we report the integration of machine learning with high-throughput experimentation to enable the rapid and large-scale identification of such nanoformulations. We identified 100 self-assembling drug nanoparticles from 2.1 million pairings, each including one of 788 candidate drugs and one of 2,686 approved excipients. We further characterized two nanoparticles, sorafenib-glycyrrhizin and terbinafine-taurocholic acid both ex vivo and in vivo. We anticipate that our platform can accelerate the development of safer and more efficacious nanoformulations with high drug-loading capacities for a wide range of therapeutics.

Allosteric Antagonist Modulation of TRPV2 by Piperlongumine Impairs Glioblastoma Progression

The use of computational tools to identify biological targets of natural products with anticancer properties and unknown modes of action is gaining momentum. We employed self-organizing maps to deconvolute the phenotypic effects of piperlongumine (PL) and establish a link to modulation of the human transient receptor potential vanilloid 2 (hTRPV2) channel. The structure of the PL-bound full-length rat TRPV2 channel was determined by cryo-EM. PL binds to a transient allosteric pocket responsible for a new mode of anticancer activity against glioblastoma (GBM) in which hTRPV2 is overexpressed. Calcium imaging experiments revealed the importance of Arg539 and Thr522 residues on the antagonistic effect of PL and calcium influx modulation of the TRPV2 channel. Downregulation of hTRPV2 reduces sensitivity to PL and decreases ROS production. Analysis of GBM patient samples associates hTRPV2 overexpression with tumor grade, disease progression, and poor prognosis. Extensive tumor abrogation and long term survival was achieved in two murine models of orthotopic GBM by formulating PL in an implantable scaffold/hydrogel for sustained local therapy. Furthermore, in primary tumor samples derived from GBM patients, we observed a selective reduction of malignant cells in response to PL ex vivo. Our results establish a broadly applicable strategy, leveraging data-motivated research hypotheses for the discovery of novel means tackling cancer.

Machine Learning Uncovers Food- and Excipient-Drug Interactions

Inactive ingredients and generally recognized as safe compounds are regarded by the US Food and Drug Administration (FDA) as benign for human consumption within specified dose ranges, but a growing body of research has revealed that many inactive ingredients might have unknown biological effects at these concentrations and might alter treatment outcomes. To speed up such discoveries, we apply state-of-the-art machine learning to delineate currently unknown biological effects of inactive ingredients—focusing on P-glycoprotein (P-gp) and uridine diphosphate-glucuronosyltransferase-2B7 (UGT2B7), two proteins that impact the pharmacokinetics of approximately 20% of FDA-approved drugs. Our platform identifies vitamin A palmitate and abietic acid as inhibitors of P-gp and UGT2B7, respectively; in silico, in vitro, ex vivo, and in vivo validations support these interactions. Our predictive framework can elucidate biological effects of commonly consumed chemical matter with implications on food-and excipient-drug interactions and functional drug formulation development.

Reker et al. use machine learning to identify biological activities of food and drug additives. Validation confirms vitamin A palmitate as an inhibitor of P-glycoprotein transport and abietic acid as an inhibitor of UGT2b7 metabolism. Such associations have important implications as food-or excipient-drug interactions.

Review on the Use of Artificial Intelligence to Predict Fire Performance of Construction Materials and Their Flame Retardancy

The evaluation and interpretation of the behavior of construction materials under fire conditions have been complicated. Over the last few years, artificial intelligence (AI) has emerged as a reliable method to tackle this engineering problem. This review summarizes existing studies that applied AI to predict the fire performance of different construction materials (e.g., concrete, steel, timber, and composites). The prediction of the flame retardancy of some structural components such as beams, columns, slabs, and connections by utilizing AI-based models is also discussed. The end of this review offers insights on the advantages, existing challenges, and recommendations for the development of AI techniques used to evaluate the fire performance of construction materials and their flame retardancy. This review offers a comprehensive overview to researchers in the fields of fire engineering and material science, and it encourages them to explore and consider the use of AI in future research projects.

Impact of sphingosine and acetylsphingosines on the aggregation and toxicity of metal-free and metal-treated amyloid-β

Pathophysiological shifts in the cerebral levels of sphingolipids in Alzheimer's disease (AD) patients suggest a link between sphingolipid metabolism and the disease pathology. Sphingosine (SP), a structural backbone of sphingolipids, is an amphiphilic molecule that is able to undergo aggregation into micelles and micellar aggregates. Considering its structural properties and cellular localization, we hypothesized that SP potentially interacts with amyloid-β (Aβ) and metal ions that are found as pathological components in AD-affected brains, with manifesting its reactivity towards metal-free Aβ and metal-bound Aβ (metal–Aβ). Herein, we report, for the first time, that SP is capable of interacting with both Aβ and metal ions and consequently affects the aggregation of metal-free Aβ and metal–Aβ. Moreover, incubation of SP with Aβ in the absence and presence of metal ions results in the aggravation of toxicity induced by metal-free Aβ and metal–Aβ in living cells. As the simplest acyl derivatives of SP, N-acetylsphingosine and 3-O-acetylsphingosine also influence metal-free Aβ and metal–Aβ aggregation to different degrees, compared to SP. Such slight structural modifications of SP neutralize its ability to exacerbate the cytotoxicity triggered by metal-free Aβ and metal–Aβ. Notably, the reactivity of SP and the acetylsphingosines towards metal-free Aβ and metal–Aβ is determined to be dependent on their formation of micelles and micellar aggregates. Our overall studies demonstrate that SP and its derivatives could directly interact with pathological factors in AD and modify their pathogenic properties at concentrations below and above critical aggregation concentrations.

The reactivity of sphingosine and acetylsphingosines towards both metal-free and metal-treated amyloid-β is demonstrated showing a correlation of their micellization properties.

Machine Learning Methods in Drug Discovery

The advancements of information technology and related processing techniques have created a fertile base for progress in many scientific fields and industries. In the fields of drug discovery and development, machine learning techniques have been used for the development of novel drug candidates. The methods for designing drug targets and novel drug discovery now routinely combine machine learning and deep learning algorithms to enhance the efficiency, efficacy, and quality of developed outputs. The generation and incorporation of big data, through technologies such as high-throughput screening and high through-put computational analysis of databases used for both lead and target discovery, has increased the reliability of the machine learning and deep learning incorporated techniques. The use of these virtual screening and encompassing online information has also been highlighted in developing lead synthesis pathways. In this review, machine learning and deep learning algorithms utilized in drug discovery and associated techniques will be discussed. The applications that produce promising results and methods will be reviewed.

Structural and biophysical insights into the mode of covalent binding of rationally designed potent BMX inhibitors

The bone marrow tyrosine kinase in chromosome X (BMX) is pursued as a drug target because of its role in various pathophysiological processes. We designed BMX covalent inhibitors with single-digit nanomolar potency with unexploited topological pharmacophore patterns. Importantly, we reveal the first X-ray crystal structure of covalently inhibited BMX at Cys496, which displays key interactions with Lys445, responsible for hampering ATP catalysis and the DFG-out-like motif, typical of an inactive conformation. Molecular dynamic simulations also showed this interaction for two ligand/BMX complexes. Kinome selectivity profiling showed that the most potent compound is the strongest binder, displays intracellular target engagement in BMX-transfected cells with two-digit nanomolar inhibitory potency, and leads to BMX degradation PC3 in cells. The new inhibitors displayed anti-proliferative effects in androgen-receptor positive prostate cancer cells that where further increased when combined with known inhibitors of related signaling pathways, such as PI3K, AKT and Androgen Receptor. We expect these findings to guide development of new selective BMX therapeutic approaches.

Mechanistic Insights into a DMSO-Perturbing Inhibitory Assay of Hyaluronidase

A hyaluronic acid-degrading enzyme (hyaluronidase; HAase) is involved in tumor growth and inflammation, and consequently, HAase inhibitors have received recent attention as potential pharmaceuticals. Previous studies have discovered a wide range of inhibitors; however, unfortunately, most of them are dissimilar to the original ligand hyaluronic acid, and their mode of inhibition remains ambiguous or seems promiscuous. This situation presents an urgent need for readily available and highly reliable assay systems identifying the promiscuous inhibitory properties of HAase inhibitors. We have previously proposed a unique method to identify promiscuous nonspecific binding inhibitors of HAase by using the DMSO-perturbing effect. Here, to obtain mechanistic insights into the DMSO-perturbing assay, we studied the addition effect of 11 water-compatible chemicals on HAase inhibitory assay. Intriguingly, the perturbing property was found to be highly specific to DMSO. Furthermore, kinetic analyses described characteristic description of the perturbing property of DMSO: DMSO displayed entropy-driven interactions with HAase, whereas nonperturbing agents such as ethanol and urea exhibited enthalpy-driven interactions. The enthalpy-driven tight interactions of ethanol and urea with HAase would lead to the irreversible denaturation of the enzymes, while the entropy-driven weak interactions caused structural and catalytic perturbation, generating nonproductive but nondenatured states of enzymes, that are key species of the perturbation assay. With these mechanistic understandings in hand, the present assay will enable rapid and reliable identification of HAase inhibitors with certain pharmaceutical potential.

SCAM Detective: Accurate Predictor of Small, Colloidally Aggregating Molecules

Small, colloidally aggregating molecules (SCAMs) are the most common source of false positives in high-throughput screening (HTS) campaigns. Although SCAMs can be experimentally detected and suppressed by the addition of detergent in the assay buffer, detergent sensitivity is not routinely monitored in HTS. Computational methods are thus needed to flag potential SCAMs during HTS triage. In this study, we have developed and rigorously validated quantitative structure-interference relationship (QSIR) models of detergent-sensitive aggregation in several HTS campaigns under various assay conditions and screening concentrations. In particular, we have modeled detergent-sensitive aggregation in an AmpC β-lactamase assay, the preferred HTS counter-screen for aggregation, as well as in another assay that measures cruzain inhibition. Our models increase the accuracy of aggregation prediction by ∼53% in the β-lactamase assay and by ∼46% in the cruzain assay compared to previously published methods. We also discuss the importance of both assay conditions and screening concentrations in the development of QSIR models for various interference mechanisms besides aggregation. The models developed in this study are publicly available for fast prediction within the SCAM detective web application (https://scamdetective.mml.unc.edu/).

The antidiabetic drug lobeglitazone has the potential to inhibit PTP1B activity

Protein tyrosine phosphatase 1B (PTP1B) is considered a potential therapeutic target for the treatment of type 2 diabetes mellitus (T2DM), since this enzyme plays a significant role to down-regulate insulin and leptin signalling and its over expression has been implicated in the development of insulin resistance, T2DM and obesity. Some thiazolidinediones (TZD) derivatives have been reported as promising PTP1B inhibitors with anti hyperglycemic effects. Recently, lobeglitazone, a new TZD, was described as an antidiabetic drug that targets the PPAR-γ (peroxisome γ proliferator-activated receptor) pathway, but no information on its effects on PTP1B have been reported to date. We investigated the effects of lobeglitazone on PTP1B activity in vitro. Surprisingly, lobeglitazone led to moderate inhibition on PTP1B (IC₅₀ 42.8 ± 3.8 µM) activity and to a non-competitive reversible mechanism of action. As lobeglitazone inhibits PTP1B activity in vitro, we speculate that it could also target PTP1B signalling pathway in vivo and thus contribute to potentiate its antidiabetic effects.

Natural products against acute respiratory infections: Strategies and lessons learned

ETHNOPHARMACOLOGICAL RELEVANCE

A wide variety of traditional herbal remedies have been used throughout history for the treatment of symptoms related to acute respiratory infections (ARIs).

AIM OF THE REVIEW

The present work provides a timely overview of natural products affecting the most common pathogens involved in ARIs, in particular influenza viruses and rhinoviruses as well as bacteria involved in co-infections, their molecular targets, their role in drug discovery, and the current portfolio of available naturally derived anti-ARI drugs.

MATERIALS AND METHODS

Literature of the last ten years was evaluated for natural products active against influenza viruses and rhinoviruses. The collected bioactive agents were further investigated for reported activities against ARI-relevant bacteria, and analysed for the chemical space they cover in relation to currently known natural products and approved drugs.

RESULTS

An overview of (i) natural compounds active in target-based and/or phenotypic assays relevant to ARIs, (ii) extracts, and (iii) in vivo data are provided, offering not only a starting point for further in-depth phytochemical and antimicrobial studies, but also revealing insights into the most relevant anti-ARI scaffolds and compound classes. Investigations of the chemical space of bioactive natural products based on principal component analysis show that many of these compounds are drug-like. However, some bioactive natural products are substantially larger and have more polar groups than most approved drugs. A workflow with various strategies for the discovery of novel antiviral agents is suggested, thereby evaluating the merit of in silico techniques, the use of complementary assays, and the relevance of ethnopharmacological knowledge on the exploration of the therapeutic potential of natural products.

CONCLUSIONS

The longstanding ethnopharmacological tradition of natural remedies against ARIs highlights their therapeutic impact and remains a highly valuable selection criterion for natural materials to be investigated in the search for novel anti-ARI acting concepts. We observe a tendency towards assaying for broad-spectrum antivirals and antibacterials mainly discovered in interdisciplinary academic settings, and ascertain a clear demand for more translational studies to strengthen efforts for the development of effective and safe therapeutic agents for patients suffering from ARIs.

The good, the bad, and the ugly in chemical and biological data for machine learning

Machine learning and artificial intelligence (ML/AI) have become important research tools in molecular medicine and chemistry. Their rise and recent success in drug discovery promises a rapid progression of development pipelines while reshaping how fundamental and clinical research is conducted. By taking advantage of the ever-growing wealth of publicly available and proprietary data, learning algorithms now provide an attractive means to generate statistically motivated research hypotheses. Hitherto unknown data patterns may guide and prioritize experiments, and augment expert intuition. Therefore, data is a key component in the model building workflow. Herein, I aim to discuss types of chemical and biological data according to their quality and reemphasize general recommendations for their use in ML/AI.

Machine learning for target discovery in drug development

The discovery of macromolecular targets for bioactive agents is currently a bottleneck for the informed design of chemical probes and drug leads. Typically, activity profiling against genetically manipulated cell lines or chemical proteomics is pursued to shed light on their biology and deconvolute drug-target networks. By taking advantage of the ever-growing wealth of publicly available bioactivity data, learning algorithms now provide an attractive means to generate statistically motivated research hypotheses and thereby prioritize biochemical screens. Here, we highlight recent successes in machine intelligence for target identification and discuss challenges and opportunities for drug discovery.

Dissecting celastrol with machine learning to unveil dark pharmacology

By coalescing bespoke machine learning and bioinformatics analyses with cell-based assays, we unveil the pharmacology of celastrol. Celastrol is a direct modulator of the progesterone and cannabinoid receptors, and its effects correlate with the antiproliferative activity. We demonstrate how in silico methods may drive systems biology studies for natural products.