Machine Learning Assisted Hit Prioritization for High Throughput Screening in Drug Discovery

Efficient prioritization of bioactive compounds from high throughput screening campaigns is a fundamental challenge for accelerating drug development efforts. In this study, we present the first data-driven approach to simultaneously detect assay interferents and prioritize true bioactive compounds. By analyzing the learning dynamics during training of a gradient boosting model on noisy high throughput screening data using a novel formulation of sample influence, we are able to distinguish between compounds exhibiting the desired biological response and those producing assay artifacts. Therefore, our method enables false positive and true positive detection without relying on prior screens or assay interference mechanisms, making it applicable to any high throughput screening campaign. We demonstrate that our approach consistently excludes assay interferents with different mechanisms and prioritizes biologically relevant compounds more efficiently than all tested baselines, including a retrospective case study simulating its use in a real drug discovery campaign. Finally, our tool is extremely computationally efficient, requiring less than 30 s per assay on low-resource hardware. As such, our findings show that our method is an ideal addition to existing false positive detection tools and can be used to guide further pharmacological optimization after high throughput screening campaigns.

MELLODDY: Cross-pharma Federated Learning at Unprecedented Scale Unlocks Benefits in QSAR without Compromising Proprietary Information

Federated multipartner machine learning has been touted as an appealing and efficient method to increase the effective training data volume and thereby the predictivity of models, particularly when the generation of training data is resource-intensive. In the landmark MELLODDY project, indeed, each of ten pharmaceutical companies realized aggregated improvements on its own classification or regression models through federated learning. To this end, they leveraged a novel implementation extending multitask learning across partners, on a platform audited for privacy and security. The experiments involved an unprecedented cross-pharma data set of 2.6+ billion confidential experimental activity data points, documenting 21+ million physical small molecules and 40+ thousand assays in on-target and secondary pharmacodynamics and pharmacokinetics. Appropriate complementary metrics were developed to evaluate the predictive performance in the federated setting. In addition to predictive performance increases in labeled space, the results point toward an extended applicability domain in federated learning. Increases in collective training data volume, including by means of auxiliary data resulting from single concentration high-throughput and imaging assays, continued to boost predictive performance, albeit with a saturating return. Markedly higher improvements were observed for the pharmacokinetics and safety panel assay-based task subsets.

AIDDISON: Empowering Drug Discovery with AI/ML and CADD Tools in a Secure, Web-Based SaaS Platform

The widespread proliferation of artificial intelligence (AI) and machine learning (ML) methods has a profound effect on the drug discovery process. However, many scientists are reluctant to utilize these powerful tools due to the steep learning curve typically associated with them. AIDDISON offers a convenient, secure, web-based platform for drug discovery, addressing the reluctance of scientists to adopt AI and ML methods due to the steep learning curve. By seamlessly integrating generative models, ADMET property predictions, searches in vast chemical spaces, and molecular docking, AIDDISON provides a sophisticated platform for modern drug discovery. It enables less computer-savvy scientists to utilize these powerful tools in their daily activities, as demonstrated by an example of identifying a valuable set of molecules for lead optimization. With AIDDISON, the benefits of AI/ML in drug discovery are accessible to all.

MEK inhibition induces expression of differentiation marker Keratin 10 in human keratinocytes

BRAF mutant metastatic melanoma was regularly treated in the past with a BRAF inhibitor (BRAFi) alone or in combination with inhibitors of the mitogen-activated protein kinase kinase (MEKi), which is still a common treatment. This combination therapy strongly reduced the occurrence of keratoacanthomas and squamous cell carcinoma, which was frequently seen when BRAFi was used as monotherapy. Here we addressed the question whether MEK inhibition counteracts squamous cell carcinoma development in part by promoting keratinocyte differentiation. Exposure of human immortalized keratinocytes to different concentrations of MEKi revealed a significant increase in the expression of differentiation-associated keratins K10 and K1 as determined by qRT-PCR and immunofluorescence staining. Taken together, the present study suggests that in a combined treatment of melanoma with BRAFi/MEKi, MEKi reduces the incidence of squamous cell carcinomas by promoting keratinocyte differentiation under combined BRAFi/MEKi treatment in melanoma. This might open further treatment perspectives for skin cancer treatment.

AI/ML combined with next-generation sequencing of VHH immune repertoires enables the rapid identification of de novo humanized and sequence-optimized single domain antibodies: a prospective case study

Introduction: In this study, we demonstrate the feasibility of yeast surface display (YSD) and nextgeneration sequencing (NGS) in combination with artificial intelligence and machine learning methods (AI/ML) for the identification of de novo humanized single domain antibodies (sdAbs) with favorable early developability profiles. Methods: The display library was derived from a novel approach, in which VHH-based CDR3 regions obtained from a llama (Lama glama), immunized against NKp46, were grafted onto a humanized VHH backbone library that was diversified in CDR1 and CDR2. Following NGS analysis of sequence pools from two rounds of fluorescence-activated cell sorting we focused on four sequence clusters based on NGS frequency and enrichment analysis as well as in silico developability assessment. For each cluster, long short-term memory (LSTM) based deep generative models were trained and used for the in silico sampling of new sequences. Sequences were subjected to sequence- and structure-based in silico developability assessment to select a set of less than 10 sequences per cluster for production. Results: As demonstrated by binding kinetics and early developability assessment, this procedure represents a general strategy for the rapid and efficient design of potent and automatically humanized sdAb hits from screening selections with favorable early developability profiles.

Metabolism of (R)-Praziquantel versus the Activation of a Parasite Transient Receptor Potential Melastatin Ion Channel

Praziquantel (PZQ) is an essential anthelmintic drug recently established to be an activator of a Transient Receptor Potential Melastatin (TRPMPZQ ) ion channel in trematode worms. Bioinformatic, mutagenesis and drug metabolism work indicate that the cyclohexyl ring of PZQ is a key pharmacophore for activation of trematode TRPMPZQ , as well as serving as the primary site of oxidative metabolism which results in PZQ being a short-lived drug. Based on our recent findings, the hydrophobic cleft in schistosome TRPMPZQ defined by three hydrophobic residues surrounding the cyclohexyl ring has little tolerance for polarity. Here we evaluate the in vitro and in vivo activities of PZQ analogues with improved metabolic stability relative to the challenge of maintaining activity on the channel. Finally, an estimation of the respective contribution to the overall activity of both the parent and the main metabolite of PZQ in humans is reported.

Practical guidelines for the use of gradient boosting for molecular property prediction

Decision tree ensembles are among the most robust, high-performing and computationally efficient machine learning approaches for quantitative structure-activity relationship (QSAR) modeling. Among them, gradient boosting has recently garnered particular attention, for its performance in data science competitions, virtual screening campaigns, and bioactivity prediction. However, different variants of gradient boosting exist, the most popular being XGBoost, LightGBM and CatBoost. Our study provides the first comprehensive comparison of these approaches for QSAR. To this end, we trained 157,590 gradient boosting models, which were evaluated on 16 datasets and 94 endpoints, comprising 1.4 million compounds in total. Our results show that XGBoost generally achieves the best predictive performance, while LightGBM requires the least training time, especially for larger datasets. In terms of feature importance, the models surprisingly rank molecular features differently, reflecting differences in regularization techniques and decision tree structures. Thus, expert knowledge must always be employed when evaluating data-driven explanations of bioactivity. Furthermore, our results show that the relevance of each hyperparameter varies greatly across datasets and that it is crucial to optimize as many hyperparameters as possible to maximize the predictive performance. In conclusion, our study provides the first set of guidelines for cheminformatics practitioners to effectively train, optimize and evaluate gradient boosting models for virtual screening and QSAR applications.

Propensity of selecting mutant parasites for the antimalarial drug cabamiquine

We report an analysis of the propensity of the antimalarial agent cabamiquine, a Plasmodium-specific eukaryotic elongation factor 2 inhibitor, to select for resistant Plasmodium falciparum parasites. Through in vitro studies of laboratory strains and clinical isolates, a humanized mouse model, and volunteer infection studies, we identified resistance-associated mutations at 11 amino acid positions. Of these, six (55%) were present in more than one infection model, indicating translatability across models. Mathematical modelling suggested that resistant mutants were likely pre-existent at the time of drug exposure across studies. Here, we estimated a wide range of frequencies of resistant mutants across the different infection models, much of which can be attributed to stochastic differences resulting from experimental design choices. Structural modelling implicates binding of cabamiquine to a shallow mRNA binding site adjacent to two of the most frequently identified resistance mutations.

Field trial: disinfection of contaminated anesthetic masks for piglets

This paper aimed to assess the success of cleaning and disinfection on microbiological contamination of anesthetic masks, which were used for automated isoflurane anesthesia for surgical castration of male piglets. Data collection took place on 11 farms in Southern Germany between September 2020 and June 2022. Each farm was visited three times (one farm having two different anesthesia devices was visited six times), and microbiological assessments took place at four sample points (SP): after unpacking the masks (SP0), after disinfection before anesthesia (SP1), after anesthesia of all piglets to be castrated in this run (SP2), and after disinfection after anesthesia (SP3). The microbiological assessment included the determination of total bacteria count, total count of hemolytic and non-hemolytic mesophilic aerotolerant bacteria and a qualitative detection of indicator bacteria Escherichia (E.) coli, extended-spectrum beta-lactamase-producing E. coli (ESBL) and methicillin-resistant Staphylococcus aureus (MRSA). For analysis, a generalized linear mixed model was applied using farms and farm visits as random effects and sampling points nested in farm visits as fixed effect. The fixed effect was highly significant for all three variables (total bacteria count, total count of hemolytic and non-hemolytic mesophilic aerotolerant bacteria) (p < 0.001). The bacterial counts at SP0 were about the same as at SP3. Concerning indicator bacteria, their presence was highest at SP2 and lowest at SP3. No indicator bacteria were present at SP1. It can be concluded that disinfection of anesthetic masks, especially before performing anesthesia, may effectively protect piglets of the following batch against unwanted transmission of pathogens. These findings will help farmers plan cleaning and disinfection activities.

Tuning gradient boosting for imbalanced bioassay modelling with custom loss functions

While in the last years there has been a dramatic increase in the number of available bioassay datasets, many of them suffer from extremely imbalanced distribution between active and inactive compounds. Thus, there is an urgent need for novel approaches to tackle class imbalance in drug discovery. Inspired by recent advances in computer vision, we investigated a panel of alternative loss functions for imbalanced classification in the context of Gradient Boosting and benchmarked them on six datasets from public and proprietary sources, for a total of 42 tasks and 2 million compounds. Our findings show that with these modifications, we achieve statistically significant improvements over the conventional cross-entropy loss function on five out of six datasets. Furthermore, by employing these bespoke loss functions we are able to push Gradient Boosting to match or outperform a wide variety of previously reported classifiers and neural networks. We also investigate the impact of changing the loss function on training time and find that it increases convergence speed up to 8 times faster. As such, these results show that tuning the loss function for Gradient Boosting is a straightforward and computationally efficient method to achieve state-of-the-art performance on imbalanced bioassay datasets without compromising on interpretability and scalability.

Graphical Abstract

Efficient secretory production of proline/alanine/serine (PAS) biopolymers in Corynebacterium glutamicum yielding a monodisperse biological alternative to polyethylene glycol (PEG)

Background

PAS biopolymers are recombinant polypeptides comprising the small uncharged l-amino acids Pro, Ala and/or Ser which resemble the widely used poly-ethylene glycol (PEG) in terms of pronounced hydrophilicity. Likewise, their random chain behaviour in physiological solution results in a strongly expanded hydrodynamic volume. Thus, apart from their use as fusion partner for biopharmaceuticals to achieve prolonged half-life in vivo, PAS biopolymers appear attractive as substitute for PEG—or other poorly degradable chemical polymers—in many areas. As a prerequisite for the wide application of PAS biopolymers at affordable cost, we have established their highly efficient biotechnological production in Corynebacterium glutamicum serving as a well characterized bacterial host organism.

Results

Using the CspA signal sequence, we have secreted two representative PAS biopolymers as polypeptides with ~ 600 and ~ 1200 amino acid residues, respectively. Both PAS biopolymers were purified from the culture supernatant by means of a simple downstream process in a truly monodisperse state as evidenced by ESI–MS. Yields after purification were up to ≥ 4 g per liter culture, with potential for further increase by strain optimization as well as fermentation and bioprocess development. Beyond direct application as hydrocolloids or to exploit their rheological properties, such PAS biopolymers are suitable for site-specific chemical conjugation with pharmacologically active molecules via their unique terminal amino or carboxyl groups. To enable the specific activation of the carboxylate, without interference by the free amino group, we generated a blocked N-terminus for the PAS(1200) polypeptide simply by introducing an N-terminal Gln residue which, after processing of the signal peptide, was cyclised to a chemically inert pyroglutamyl group upon acid treatment. The fact that PAS biopolymers are genetically encoded offers further conjugation strategies via incorporation of amino acids with reactive side chains (e.g., Cys, Lys, Glu/Asp) at defined positions.

Conclusions

Our new PAS expression platform using Corynex^® technology opens the way to applications of PASylation^® technology in multiple areas such as the pharmaceutical industry, cosmetics and food technology.

Effect of ascorbic acid and acerola juice on some quality properties of aseptic filled apple juice

Quality and storability are crucial factors in production of apple juice. The main goal of this study was investigation of the effects of ascorbic acid and acerola juice on the changes of some sensorial parameters and bioactive compounds of aseptically filled and industrial scale produced apple juice during storage for 12 months. While the viscosity and pH of apple juice did not change significantly, the ΔE* peaked (20–30) at month 6 of the storage period. The colour of apple juice was lighter than at the beginning of storage. Maximum total phenolic contents were 1,100, 1,400, and 1,250 mg L−1 in the control, ascorbic acid supplemented, and acerola added samples, respectively. Other parameters (antioxidant capacity, ascorbic acid, browning index, etc.) peaked in month 4. Acerola was a good alternative anti-browning and antioxidant agent for the treatment of apple juice in the processing. The antioxidant capacity of apple juice treated with acerola was higher than with ascorbic acid. The results were obtained with industrial samples, thus, they can serve as a very good base for the optimisation process and industrial production without the need for scale-up.

Mechanism of praziquantel action at a parasitic flatworm ion channel

[Figure: see text].

Learning from Nature: From a Marine Natural Product to Synthetic Cyclooxygenase-1 Inhibitors by Automated De Novo Design

The repertoire of natural products offers tremendous opportunities for chemical biology and drug discovery. Natural product-inspired synthetic molecules represent an ecologically and economically sustainable alternative to the direct utilization of natural products. De novo design with machine intelligence bridges the gap between the worlds of bioactive natural products and synthetic molecules. On employing the compound Marinopyrrole A from marine Streptomyces as a design template, the algorithm constructs innovative small molecules that can be synthesized in three steps, following the computationally suggested synthesis route. Computational activity prediction reveals cyclooxygenase (COX) as a putative target of both Marinopyrrole A and the de novo designs. The molecular designs are experimentally confirmed as selective COX-1 inhibitors with nanomolar potency. X-ray structure analysis reveals the binding of the most selective compound to COX-1. This molecular design approach provides a blueprint for natural product-inspired hit and lead identification for drug discovery with machine intelligence.

Physicochemical and microbiological changes of bovine blood due to high hydrostatic pressure treatment

Bovine blood samples were treated with high hydrostatic pressure (HHP) to examine the changes that may occur in the blood related to its colour, microbiological characteristics, protein denaturation, and dynamic viscosity. Pressure treatments were carried out from 100 to 600 MPa in 100 MPa scale up, with 5 min holding time. The blood samples were treated with anticoagulant (EDTA) to eliminate the possible measurement distorting effects. We found that 2 log reduction in the microbial load could be achieved with a pressure treatment above 400 MPa. According to the protein denaturation measurements (DSC), blood proteins were resistant to pressure treatment, even at 300–400 MPa a substantial part of proteins remained in native state. The colour of the samples got darker with the rising pressure, however, visible colour change was observed only above 400 MPa. It can be established, that the HHP treatment was suitable to increase the microbiological stability of blood, without significantly changing its techno-functional properties.

Interobserver reliability of measures of the Welfare Quality® animal welfare assessment protocol for sows and piglets

The aim of this study was to assess the interobserver reliability of the measures forming the Welfare Quality® animal welfare assessment protocol for sows and piglets. The study was carried out at nine farms in Northern Germany. Two trained observers evaluated identical animals simultaneously but independently in 40 joint farm visits. Interobserver reliability was calculated at individual animal level using Cohen's kappa, weighted kappa and the prevalence-adjusted, bias-adjusted kappa (PABAK) and at farm level using Spearman's rank correlation coefficient (RS), the intraclass correlation coefficient (ICC), smallest detectable change (SDC) and limits of agreement (LoA). While a direct comparison of the adjectives of the qualitative behaviour assessment showed poor interobserver reliability, a Principal Component Analysis detected good interobserver reliability. The assessment of social and exploratory behaviours showed acceptable interobserver reliability, while the assessment of stereotypies displayed good interobserver reliability. The human-animal relationship test showed only poor interobserver reliability at individual animal and farm levels. In most cases, measures of health and physical state assessed in sows and piglets exhibited acceptable or good interobserver reliability. In conclusion, after some measures are revised, particularly those examining the human-animal relationship, the Welfare Quality® protocol for sows and piglets will represent a reliable approach in terms of interobserver reliability to assess the welfare of sows and piglets.

Shape Similarity by Fractal Dimensionality: An Application in the de novo Design of (-)-Englerin A Mimetics

Molecular shape and pharmacological function are interconnected. To capture shape, the fractal dimensionality concept was employed, providing a natural similarity measure for the virtual screening of de novo generated small molecules mimicking the structurally complex natural product (-)-englerin A. Two of the top-ranking designs were synthesized and tested for their ability to modulate transient receptor potential (TRP) cation channels which are cellular targets of (-)-englerin A. Intracellular calcium assays and electrophysiological whole-cell measurements of TRPC4 and TRPM8 channels revealed potent inhibitory effects of one of the computer-generated compounds. Four derivatives of this identified hit compound had comparable effects on TRPC4 and TRPM8. The results of this study corroborate the use of fractal dimensionality as an innovative shape-based molecular representation for molecular scaffold-hopping.

Design of Natural-Product-Inspired Multitarget Ligands by Machine Learning

A virtual screening protocol based on machine learning models was used to identify mimetics of the natural product (-)-galantamine. This fully automated approach identified eight compounds with bioactivities on at least one of the macromolecular targets of (-)-galantamine, with different polypharmacological profiles. Two of the computer-generated hits possess an expanded spectrum of bioactivity on targets relevant to the treatment of Alzheimer's disease and are suitable for hit-to-lead expansion. These results advocate multitarget drug design by advanced virtual screening protocols based on chemically informed machine learning models.

Discovery of Novel Molecular Frameworks of Farnesoid X Receptor Modulators by Ensemble Machine Learning

The bile acid activated transcription factor farnesoid X receptor (FXR) has revealed therapeutic potential as a molecular drug target for the treatment of hepatic and metabolic disorders. Despite strong efforts in FXR ligand development, the structural diversity among the known FXR modulators is limited. Only four molecular frameworks account for more than 50 % of the FXR modulators annotated in ChEMBL. Here, we leverage machine learning methods to expand the chemical space of FXR-targeting small molecules by employing an ensemble of three complementary machine learning approaches. A counter-propagation artificial neural network, a k-nearest neighbor learner, and a three-dimensional pharmacophore descriptor were combined to retrieve novel FXR ligands from a collection of more than 3 million compounds. The ensemble machine learning model identified six new FXR modulators among ten top-ranked candidates. These active hits comprise both FXR activators and antagonists with micromolar potencies. With four novel FXR ligand scaffolds, these computationally identified bioactive compounds appreciably expand the chemical space of known FXR modulators and may serve as starting points for hit-to-lead expansion.

Discovery of Novel Molecular Frameworks of Farnesoid X Receptor Modulators by Ensemble Machine Learning

Invited for this month's cover picture is the group of Prof. Dr. Gisbert Schneider from the Swiss Federal Institute of Technology (ETH) Zurich (Switzerland). The cover picture illustrates the application of machine-learning methods to expand the chemical space of farnesoid X receptor (FXR)-targeting small molecules, by employing an ensemble of three complementary machine-learning approaches (counter-propagation artificial neural network, k-nearest neighbor learner, and three-dimensional pharmacophore model). Read the full text of their Full Paper at 10.1002/open.201800156.

Front Cover: Discovery of Novel Molecular Frameworks of Farnesoid X Receptor Modulators by Ensemble Machine Learning (ChemistryOpen 1/2019)

The Front Cover shows the application of machine‐learning methods to expand the chemical space of farnesoid X receptor (FXR)‐targeting small molecules, by employing an ensemble of three complementary machine‐learning approaches (counter‐propagation artificial neural network, k‐nearest neighbor learner, and three‐dimensional pharmacophore model). The ensemble machine‐learning model identified six new FXR modulators from a library of 3 million compounds. These computationally identified bioactive compounds possess four novel scaffolds and appreciably expand the chemical space of known FXR modulators. More information can be found in the Full Paper by D. Merk et al. on page 7 in Issue 1, 2019 (DOI: 10.1002/open.201800156).

Scaffold hopping from synthetic RXR modulators by virtual screening and de novo design

The lack of potent subtype-selective modulators of retinoid X receptors (RXRs) has hindered their full exploitation as promising drug targets. Using computational similarity searching, target prediction and automated de novo design, we identified novel RXR ligands exhibiting innovative molecular frameworks, pronounced receptor-subtype preference and suitable properties for hit-to-lead expansion.

De Novo Design of Bioactive Small Molecules by Artificial Intelligence

Generative artificial intelligence offers a fresh view on molecular design. We present the first-time prospective application of a deep learning model for designing new druglike compounds with desired activities. For this purpose, we trained a recurrent neural network to capture the constitution of a large set of known bioactive compounds represented as SMILES strings. By transfer learning, this general model was fine-tuned on recognizing retinoid X and peroxisome proliferator-activated receptor agonists. We synthesized five top-ranking compounds designed by the generative model. Four of the compounds revealed nanomolar to low-micromolar receptor modulatory activity in cell-based assays. Apparently, the computational model intrinsically captured relevant chemical and biological knowledge without the need for explicit rules. The results of this study advocate generative artificial intelligence for prospective de novo molecular design, and demonstrate the potential of these methods for future medicinal chemistry.

[Complications of body piercing]

The trend of body piercing has grown in popularity in the past decade within the general population and especially among young adults. Complications of body piercing include local inflammation and infections, but severe complications are also possible and largely underestimated. People are usually not aware of the risks before making a piercing, and their medical history, medication and comorbidities are largely neglected by the people who realise the piercing. This article presents a review of the complications that a primary care physician may observe, for a patient who wishes to make a piercing, or presents complications due to the implementation of such a device.

A central role of salicylic Acid in plant disease resistance

Transgenic tobacco and Arabidopsis thaliana expressing the bacterial enzyme salicylate hydroxylase cannot accumulate salicylic acid (SA). This defect not only makes the plants unable to induce systemic acquired resistance, but also leads to increased susceptibility to viral, fungal, and bacterial pathogens. The enhanced susceptibility extends even to host-pathogen combinations that would normally result in genetic resistance. Therefore, SA accumulation is essential for expression of multiple modes of plant disease resistance.

Requirement of salicylic Acid for the induction of systemic acquired resistance

It has been proposed that salicylic acid acts as an endogenous signal responsible for inducing systemic acquired resistance in plants. The contribution of salicylic acid to systemic acquired resistance was investigated in transgenic tobacco plants harboring a bacterial gene encoding salicylate hydroxylase, which converts salicylic acid to catechol. Transgenic plants that express salicylate hydroxylase accumulated little or no salicylic acid and were defective in their ability to induce acquired resistance against tobacco mosaic virus. Thus, salicylic acid is essential for the development of systemic acquired resistance in tobacco.