Optimal HTS Fingerprint Definitions by Using a Desirability Function and a Genetic Algorithm.

The use of compound biological fingerprints built on data from high-throughput screening (HTS) campaigns, or HTS fingerprints, is a novel cheminformatics method of representing compounds by integrating chemical and biological activity data that is gaining momentum in its application to drug discovery, including hit expansion, target identification, and virtual screening. HTS fingerprints present two major limitations, noise and missing data, which are intrinsic to the high-throughput data acquisition technologies and to the assay availability or assay selection procedure used for their construction. In this work, we present a methodology to define an optimal set of HTS fingerprints by using a desirability function that encodes the principles of maximum biological and chemical space coverage and minimum redundancy between HTS assays. We used a genetic algorithm to optimize the desirability function and obtained an optimal fingerprint that was evaluated for performance in a test set of 33 diverse assays. Our results show that the optimal HTS fingerprint represents compounds in chemical biology space using 25% fewer assays. When used for virtual screening, the optimal HTS fingerprint obtained equivalent performance, in terms of both area under the curve and enrichment factors, to full fingerprints for 27 out of 33 test assays, while randomly assembled fingerpints could achieve equivalent performance in only 23 test assays.

Machine learning approach to identify malaria risk in travelers using real-world evidence.

Background: Pre-travel consultation and chemoprophylaxis measures for malaria are a key component in the prevention of imported malaria in travelers. In this study we report a predictive tool for assessing personalized malaria risk in travelers based on the analysis of electronic medical records from travel consultations. The tool aims to guide physicians in the recommendation of appropriate prophylaxis prior to their trip. We also provide best-practice recommendations for pre-processing noisy and highly sparse real world evidence data. Methods: We leveraged a large EMR dataset, containing demographic information about travelers and their destination. The data has been previously preprocessed using various strategies to handle missing and unbalanced data. We compared multiple machine learning approaches to assess the risk of malaria acquisition in travelers during their travels. Additionally, a feature importance analysis was performed using SHAP (SHapley Additive Explanations) values to identify patterns associated with malaria risk. Results: Our study revealed that our XGB models achieved high predictive capacity (AUC >0.80). The most significant features predicting malaria infection during travel included travel destinations with low malaria risk, vaccination history, number of countries visited, age, and trip duration. Remarkably, we were able to obtain a reduced model with only five features. When comparing this model with a population of travelers recommended for malaria chemoprophylaxis, we observed that it was deemed necessary in only 40% of these travelers. This suggests that 60% received chemoprophylaxis despite having a low personalized risk of malaria. Conclusion: We have developed an algorithmic tool that utilizes a concise survey to generate a personalized travel risk assessment, effectively minimizing the prescription of unnecessary malaria chemoprophylaxis. Through the identification of patterns linked to predictions, our model significantly enhances the efficacy of pre-travel consultations.

MALrisk: a machine-learning­based tool to predict imported malaria in returned travellers with fever

BACKGROUND: Early diagnosis is key to reducing the morbi-mortality associated with P. falciparum malaria among international travellers. However, access to microbiological tests can be challenging for some healthcare settings. Artificial Intelligence could improve the management of febrile travellers. METHODS: Data from a multicentric prospective study of febrile travellers was obtained to build a machine-learning model to predict malaria cases among travellers presenting with fever. Demographic characteristics, clinical and laboratory variables were leveraged as features. Eleven machine-learning classification models were evaluated by 50-fold cross-validation in a Training set. Then, the model with the best performance, defined by the Area Under the Curve (AUC), was chosen for parameter optimization and evaluation in the Test set. Finally, a reduced model was elaborated with those features that contributed most to the model. RESULTS: Out of eleven machine-learning models, XGBoost presented the best performance (mean AUC of 0.98 and a mean F1 score of 0.78). A reduced model (MALrisk) was developed using only six features: Africa as a travel destination, platelet count, rash, respiratory symptoms, hyperbilirubinemia and chemoprophylaxis intake. MALrisk predicted malaria cases with 100% (95%CI 96-100) sensitivity and 72% (95%CI 68-75) specificity. CONCLUSIONS: The MALrisk can aid in the timely identification of malaria in non-endemic settings, allowing the initiation of empiric antimalarials and reinforcing the need for urgent transfer in healthcare facilities with no access to malaria diagnostic tests. This resource could be easily scalable to a digital application and could reduce the morbidity associated with late diagnosis.

The impact of urban configuration types on urban heat islands, air pollution, CO2 emissions, and mortality in Europe: a data science approach.

BACKGROUND: The world is becoming increasingly urbanised. As cities around the world continue to grow, it is important for urban planners and policy makers to understand how different urban configuration patterns affect the environment and human health. However, previous studies have provided mixed findings. We aimed to identify European urban configuration types, on the basis of the local climate zones categories and street design variables from Open Street Map, and evaluate their association with motorised traffic flows, surface urban heat island (SUHI) intensities, tropospheric NO2, CO2 per person emissions, and age-standardised mortality. METHODS: We considered 946 European cities from 31 countries for the analysis defined in the 2018 Urban Audit database, of which 919 European cities were analysed. Data were collected at a 250 m × 250 m grid cell resolution. We divided all cities into five concentric rings based on the Burgess concentric urban planning model and calculated the mean values of all variables for each ring. First, to identify distinct urban configuration types, we applied the Uniform Manifold Approximation and Projection for Dimension Reduction method, followed by the k-means clustering algorithm. Next, statistical differences in exposures (including SUHI) and mortality between the resulting urban configuration types were evaluated using a Kruskal-Wallis test followed by a post-hoc Dunn's test. FINDINGS: We identified four distinct urban configuration types characterising European cities: compact high density (n=246), open low-rise medium density (n=245), open low-rise low density (n=261), and green low density (n=167). Compact high density cities were a small size, had high population densities, and a low availability of natural areas. In contrast, green low density cities were a large size, had low population densities, and a high availability of natural areas and cycleways. The open low-rise medium and low density cities were a small to medium size with medium to low population densities and low to moderate availability of green areas. Motorised traffic flows and NO2 exposure were significantly higher in compact high density and open low-rise medium density cities when compared with green low density and open low-rise low density cities. Additionally, green low density cities had a significantly lower SUHI effect compared with all other urban configuration types. Per person CO2 emissions were significantly lower in compact high density cities compared with green low density cities. Lastly, green low density cities had significantly lower mortality rates when compared with all other urban configuration types. INTERPRETATION: Our findings indicate that, although the compact city model is more sustainable, European compact cities still face challenges related to poor environmental quality and health. Our results have notable implications for urban and transport planning policies in Europe and contribute to the ongoing discussion on which city models can bring the greatest benefits for the environment, climate, and health. FUNDING: Spanish Ministry of Science and Innovation, State Research Agency, Generalitat de Catalunya, Centro de Investigación Biomédica en red Epidemiología y Salud Pública, and Urban Burden of Disease Estimation for Policy Making as a Horizon Europe project.

Gut microbiota composition in travellers is associated with faecal lipocalin-2, a mediator of gut inflammation.

Introduction: We examined the gut microbiota of travellers returning from tropical areas with and without traveller's diarrhoea (TD) and its association with faecal lipocalin-2 (LCN2) levels. Methods: Participants were recruited at the Hospital Clinic of Barcelona, Spain, and a single stool sample was collected from each individual to perform the diagnostic of the etiological agent causing gastrointestinal symptoms as well as to measure levels of faecal LCN2 as a biomarker of gut inflammation. We also characterised the composition of the gut microbiota by sequencing the region V3-V4 from the 16S rRNA gene, and assessed its relation with the clinical presentation of TD and LCN2 levels using a combination of conventional statistical tests and unsupervised machine learning approaches. Results: Among 61 participants, 45 had TD, with 40% having identifiable etiological agents. Surprisingly, LCN2 levels were similar across groups, suggesting gut inflammation occurs without clinical TD symptoms. Differential abundance (DA) testing highlighted a microbial profile tied to high LCN2 levels, marked by increased Proteobacteria and Escherichia-Shigella, and decreased Firmicutes, notably Oscillospiraceae. UMAP analysis confirmed this profile's association, revealing distinct clusters based on LCN2 levels. The study underscores the discriminatory power of UMAP in capturing meaningful microbial patterns related to clinical variables. No relevant differences in the gut microbiota composition were found between travellers with or without TD. Discussion: The findings suggest a correlation between gut microbiome and LCN2 levels during travel, emphasising the need for further research to discern the nature of this relationship.

Sociodemographic, lifestyle, and environmental determinants of vitamin D levels in pregnant women in Spain.

INTRODUCTION: Vitamin D deficiency (<20 ng/mL circulating levels) is a worldwide public health concern and pregnant women are especially vulnerable, affecting the health of the mother and the fetus. This study aims to evaluate the sociodemographic, lifestyle, and environmental determinants associated with circulating vitamin D levels in Spanish pregnant women. METHODS: We used data from the Spanish INMA ("Infancia y Medio Ambiente") prospective birth cohort study from the regions of Gipuzkoa, Sabadell, and Valencia. 25-hydroxyvitamin D3 (25(OH)D3) was measured in plasma collected in the first trimester of pregnancy. Information on 108 determinants was gathered: 13 sociodemographic, 48 lifestyle including diet, smoking and physical activity, and 47 environmental variables, representing the urban and the chemical exposome. Association of the determinants with maternal 25(OH)D3 levels was estimated in single- and multiple-exposure models. Machine learning techniques were used to predict 25(OH)D3 levels below sufficiency (30 ng/mL). RESULTS: The prevalence of < 30 ng/mL 25(OH)D3 levels was 51 %. In the single-exposure analysis, older age, higher socioeconomic status, taking vitamin D, B12 and other sup*plementation, and higher humidity, atmospheric pressure and UV rays were associated with higher levels of 25(OH)D3 (IQR increase of age: 1.2 [95 % CI: 0.6, 1.8] ng/mL 25(OH)D3). In the multiple-exposures model, most of these associations remained and others were revealed. Higher body mass index, PM2.5 and high deprivation area were associated with lower 25(OH)D3 levels (i.e., Quartile 4 of PM2.5 vs Q1: -3.6 [95 % CI: -5.6, -1.5] ng/mL of 25(OH)D3). History of allergy and asthma, being multiparous, intake of vegetable fat, vitamin B6, alcohol consumption and molybdenum were associated with higher levels. The machine learning classification model confirmed some of these associations. CONCLUSIONS: This comprehensive study shows that younger age, higher body mass index, higher deprived areas, higher air pollution and lower UV rays and humidity are associated with lower 25(OH)D3 levels.

Chemotography for multi-target SAR analysis in the context of biological pathways.

The increasing amount of chemogenomics data, that is, activity measurements of many compounds across a variety of biological targets, allows for better understanding of pharmacology in a broad biological context. Rather than assessing activity at individual biological targets, today understanding of compound interaction with complex biological systems and molecular pathways is often sought in phenotypic screens. This perspective poses novel challenges to structure-activity relationship (SAR) assessment. Today, the bottleneck of drug discovery lies in the understanding of SAR of rich datasets that go beyond single targets in the context of biological pathways, potential off-targets, and complex selectivity profiles. To aid in the understanding and interpretation of such complex SAR, we introduce Chemotography (chemotype chromatography), which encodes chemical space using a color spectrum by combining clustering and multidimensional scaling. Rich biological data in our approach were visualized using spatial dimensions traditionally reserved for chemical space. This allowed us to analyze SAR in the context of target hierarchies and phylogenetic trees, two-target activity scatter plots, and biological pathways. Chemotography, in combination with the Kyoto Encyclopedia of Genes and Genomes (KEGG), also allowed us to extract pathway-relevant SAR from the ChEMBL database. We identified chemotypes showing polypharmacology and selectivity-conferring scaffolds, even in cases where individual compounds have not been tested against all relevant targets. In addition, we analyzed SAR in ChEMBL across the entire Kinome, going beyond individual compounds. Our method combines the strengths of chemical space visualization for SAR analysis and graphical representation of complex biological data. Chemotography is a new paradigm for chemogenomic data visualization and its versatile applications presented here may allow for improved assessment of SAR in biological context, such as phenotypic assay hit lists.

State-of-the-art methods for exposure-health studies: Results from the exposome data challenge event.

The exposome recognizes that individuals are exposed simultaneously to a multitude of different environmental factors and takes a holistic approach to the discovery of etiological factors for disease. However, challenges arise when trying to quantify the health effects of complex exposure mixtures. Analytical challenges include dealing with high dimensionality, studying the combined effects of these exposures and their interactions, integrating causal pathways, and integrating high-throughput omics layers. To tackle these challenges, the Barcelona Institute for Global Health (ISGlobal) held a data challenge event open to researchers from all over the world and from all expertises. Analysts had a chance to compete and apply state-of-the-art methods on a common partially simulated exposome dataset (based on real case data from the HELIX project) with multiple correlated exposure variables (P > 100 exposure variables) arising from general and personal environments at different time points, biological molecular data (multi-omics: DNA methylation, gene expression, proteins, metabolomics) and multiple clinical phenotypes in 1301 mother-child pairs. Most of the methods presented included feature selection or feature reduction to deal with the high dimensionality of the exposome dataset. Several approaches explicitly searched for combined effects of exposures and/or their interactions using linear index models or response surface methods, including Bayesian methods. Other methods dealt with the multi-omics dataset in mediation analyses using multiple-step approaches. Here we discuss features of the statistical models used and provide the data and codes used, so that analysts have examples of implementation and can learn how to use these methods. Overall, the exposome data challenge presented a unique opportunity for researchers from different disciplines to create and share state-of-the-art analytical methods, setting a new standard for open science in the exposome and environmental health field.

Strong off-target antibody reactivity to malarial antigens induced by RTS,S/AS01E vaccination is associated with protection.

The RTS,S/AS01E vaccine targets the circumsporozoite protein (CSP) of the Plasmodium falciparum (P. falciparum) parasite. Protein microarrays were used to measure levels of IgG against 1000 P. falciparum antigens in 2138 infants (age 6-12 weeks) and children (age 5-17 months) from 6 African sites of the phase III trial, sampled before and at 4 longitudinal visits after vaccination. One month postvaccination, IgG responses to 17% of all probed antigens showed differences between RTS,S/AS01E and comparator vaccination groups, whereas no prevaccination differences were found. A small subset of antigens presented IgG levels reaching 4- to 8-fold increases in the RTS,S/AS01E group, comparable in magnitude to anti-CSP IgG levels (~11-fold increase). They were strongly cross-correlated and correlated with anti-CSP levels, waning similarly over time and reincreasing with the booster dose. Such an intriguing phenomenon may be due to cross-reactivity of anti-CSP antibodies with these antigens. RTS,S/AS01E vaccinees with strong off-target IgG responses had an estimated lower clinical malaria incidence after adjusting for age group, site, and postvaccination anti-CSP levels. RTS,S/AS01E-induced IgG may bind strongly not only to CSP, but also to unrelated malaria antigens, and this seems to either confer, or at least be a marker of, increased protection from clinical malaria.

Side-chain recognition and gating in the ribosome exit tunnel.

The ribosome is a large complex catalyst responsible for the synthesis of new proteins, an essential function for life. New proteins emerge from the ribosome through an exit tunnel as nascent polypeptide chains. Recent findings indicate that tunnel interactions with the nascent polypeptide chain might be relevant for the regulation of translation. However, the specific ribosomal structural features that mediate this process are unknown. Performing molecular dynamics simulations, we are studying the interactions between components of the ribosome exit tunnel and different chemical probes (specifically different amino acid side chains or monovalent inorganic ions). Our free-energy maps describe the physicochemical environment of the tunnel, revealing binding crevices and free-energy barriers for single amino acids and ions. Our simulations indicate that transport out of the tunnel could be different for diverse amino acid species. In addition, our results predict a notable protein-RNA interaction between a flexible 23S rRNA tetraloop (gate) and ribosomal protein L39 (latch) that could potentially obstruct the tunnel's exit. By relating our simulation data to earlier biochemical studies, we propose that ribosomal features at the exit of the tunnel can play a role in the regulation of nascent chain exit and ion flux. Moreover, our free-energy maps may provide a context for interpreting sequence-dependent nascent chain phenomenology.

Projection to Latent Spaces Disentangles Pathological Effects on Brain Morphology in the Asymptomatic Phase of Alzheimer's Disease.

Alzheimer's disease (AD) continuum is defined as a cascade of several neuropathological processes that can be measured using biomarkers, such as cerebrospinal fluid (CSF) levels of Aß, p-tau, and t-tau. In parallel, brain anatomy can be characterized through imaging techniques, such as magnetic resonance imaging (MRI). In this work we relate both sets of measurements and seek associations between biomarkers and the brain structure that can be indicative of AD progression. The goal is to uncover underlying multivariate effects of AD pathology on regional brain morphological information. For this purpose, we used the projection to latent structures (PLS) method. Using PLS, we found a low dimensional latent space that best describes the covariance between both sets of measurements on the same subjects. Possible confounder effects (age and sex) on brain morphology are included in the model and regressed out using an orthogonal PLS model. We looked for statistically significant correlations between brain morphology and CSF biomarkers that explain part of the volumetric variance at each region-of-interest (ROI). Furthermore, we used a clustering technique to discover a small set of CSF-related patterns describing the AD continuum. We applied this technique to the study of subjects in the whole AD continuum, from the pre-clinical asymptomatic stages all the way through to the symptomatic groups. Subsequent analyses involved splitting the course of the disease into diagnostic categories: cognitively unimpaired subjects (CU), mild cognitively impaired subjects (MCI), and subjects with dementia (AD-dementia), where all symptoms were due to AD.

Shared Latent Structures Between Imaging Features and Biomarkers in Early Stages of Alzheimer's Disease: A Predictive Study.

Magnetic resonance imaging (MRI) provides high resolution brain morphological information and is used as a biomarker in neurodegenerative diseases. Population studies of brain morphology often seek to identify pathological structural changes related to different diagnostic categories (e.g.: controls, mild cognitive impairment or dementia) which normally describe highly heterogeneous groups with a single categorical variable. Instead, multiple biomarkers are used as a proxy for pathology and are more powerful in capturing structural variability. Hence, using the joint modeling of brain morphology and biomarkers, we aim at describing structural changes related to any brain condition by means of few underlying processes. In this regard, we use a multivariate approach based on Projection to Latent Structures in its regression variant (PLSR) to study structural changes related to aging and AD pathology. MRI volumetric and cortical thickness measurements are used for brain morphology and cerebrospinal fluid (CSF) biomarkers (t-tau, p-tau and amyloid-beta) are used as a proxy for AD pathology. By relating both sets of measurements, PLSR finds a low-dimensional latent space describing AD pathological effects on brain structure. The proposed framework allows us to separately model aging effects on brain morphology as a confounder variable orthogonal to the pathological effect. The predictive power of the associated latent spaces (i.e., the capacity of predicting biomarker values) is assessed in a cross-validation framework.

Identification of NEK3 and MOK as novel targets for lithium.

Lithium ion, commonly used as the carbonate salt in the treatment of bipolar disorders, has been identified as an inhibitor of several kinases, including Glycogen Synthase Kinase-3ß, for almost 20 years. However, both the exact mechanism of enzymatic inhibition and its apparent specificity for certain metalloenzymes are still a matter of debate. A data-driven hypothesis is presented that accounts for the specificity profile of kinase inhibition by lithium in terms of the presence of a unique protein environment in the magnesium-binding site. This hypothesis has been validated by the discovery of two novel potential targets for lithium, namely NEK3 and MOK, which are related to neuronal function.

Prediction of amyloid pathology in cognitively unimpaired individuals using voxel-wise analysis of longitudinal structural brain MRI.

BACKGROUND: Magnetic resonance imaging (MRI) has unveiled specific alterations at different stages of Alzheimer's disease (AD) pathophysiologic continuum constituting what has been established as "AD signature". To what extent MRI can detect amyloid-related cerebral changes from structural MRI in cognitively unimpaired individuals is still an area open for exploration. METHOD: Longitudinal 3D-T1 MRI scans were acquired from a subset of the ADNI cohort comprising 403 subjects: 79 controls (Ctrls), 50 preclinical AD (PreAD), and 274 MCI and dementia due to AD (MCI/AD). Amyloid CSF was used as gold-standard measure with established cutoffs (< 192 pg/mL) to establish diagnostic categories. Cognitively unimpaired individuals were defined as Ctrls if were amyloid negative and PreAD otherwise. The MCI/AD group was amyloid positive. Only subjects with the same diagnostic category at baseline and follow-up visits were considered for the study. Longitudinal morphometric analysis was performed using SPM12 to calculate Jacobian determinant maps. Statistical analysis was carried out on these Jacobian maps to identify structural changes that were significantly different between diagnostic categories. A machine learning classifier was applied on Jacobian determinant maps to predict the presence of abnormal amyloid levels in cognitively unimpaired individuals. The performance of this classifier was evaluated using receiver operating characteristic curve analysis and as a function of the follow-up time between MRI scans. We applied a cost function to assess the benefit of using this classifier in the triaging of individuals in a clinical trial-recruitment setting. RESULTS: The optimal follow-up time for classification of Ctrls vs PreAD was Δt > 2.5 years, and hence, only subjects within this temporal span are used for evaluation (15 Ctrls, 10 PreAD). The longitudinal voxel-based classifier achieved an AUC = 0.87 (95%CI 0.72-0.97). The brain regions that showed the highest discriminative power to detect amyloid abnormalities were the medial, inferior, and lateral temporal lobes; precuneus; caudate heads; basal forebrain; and lateral ventricles. CONCLUSIONS: Our work supports that machine learning applied to longitudinal brain volumetric changes can be used to predict, with high precision, the presence of amyloid abnormalities in cognitively unimpaired subjects. Used as a triaging method to identify a fixed number of amyloid-positive individuals, this longitudinal voxel-wise classifier is expected to avoid 55% of unnecessary CSF and/or PET scans and reduce economic cost by 40%.

MRI-Based Screening of Preclinical Alzheimer's Disease for Prevention Clinical Trials.

The identification of healthy individuals harboring amyloid pathology represents one important challenge for secondary prevention clinical trials in Alzheimer's disease (AD). Consequently, noninvasive and cost-efficient techniques to detect preclinical AD constitute an unmet need of critical importance. In this manuscript, we apply machine learning to structural MRI (T1 and DTI) of 96 cognitively normal subjects to identify amyloid-positive ones. Models were trained on public ADNI data and validated on an independent local cohort. Used for subject classification in a simulated clinical trial setting, the proposed method is able to save 60% of unnecessary CSF/PET tests and to reduce 47% of the cost of recruitment. This recruitment strategy capitalizes on available MR scans to reduce the overall amount of invasive PET/CSF tests in prevention trials, demonstrating a potential value as a tool for preclinical AD screening. This protocol could foster the development of secondary prevention strategies for AD.

HtrA1 activation is driven by an allosteric mechanism of inter-monomer communication.

The human protease family HtrA is responsible for preventing protein misfolding and mislocalization, and a key player in several cellular processes. Among these, HtrA1 is implicated in several cancers, cerebrovascular disease and age-related macular degeneration. Currently, HtrA1 activation is not fully characterized and relevant for drug-targeting this protease. Our work provides a mechanistic step-by-step description of HtrA1 activation and regulation. We report that the HtrA1 trimer is regulated by an allosteric mechanism by which monomers relay the activation signal to each other, in a PDZ-domain independent fashion. Notably, we show that inhibitor binding is precluded if HtrA1 monomers cannot communicate with each other. Our study establishes how HtrA1 trimerization plays a fundamental role in proteolytic activity. Moreover, it offers a structural explanation for HtrA1-defective pathologies as well as mechanistic insights into the degradation of complex extracellular fibrils such as tubulin, amyloid beta and tau that belong to the repertoire of HtrA1.

Aggregated Compound Biological Signatures Facilitate Phenotypic Drug Discovery and Target Elucidation.

Predicting the cellular response of compounds is a challenge central to the discovery of new drugs. Compound biological signatures have risen as a way of representing the perturbation produced by a compound in the cell. However, their ability to encode specific phenotypic information and generating tangible predictions remains unknown, mainly because of the inherent noise in such data sets. In this work, we statistically aggregate signals from several compound biological signatures to find compounds that produce a desired phenotype in the cell. We exploit this method in two applications relevant for phenotypic screening in drug discovery programs: target-independent hit expansion and target identification. As a result, we present here (i) novel nanomolar inhibitors of cellular division that reproduce the phenotype and the mode of action of reference natural products and (ii) blockers of the NKCC1 cotransporter for autism spectrum disorders. Our results were confirmed in both cellular and biochemical assays of the respective projects. In addition, these examples provided novel insights on the information content and biological significance of compound biological signatures from HTS, and their applicability to drug discovery in general. For target identification, we show that novel targets can be predicted successfully for drugs by reporting new activities for nimedipine, fluspirilene, and pimozide and providing a rationale for repurposing and side effects. Our results highlight the opportunities of reusing public bioactivity data for prospective drug discovery, including scenarios where the effective target or mode of action of a particular molecule is not known, such as in phenotypic screening campaigns.

MHC-peptide binding is assisted by bound water molecules.

Water plays an important role in determining the high affinity of epitopes to the class I MHC complex. To study the energy and dynamics of water interactions in the complex we performed molecular dynamics simulation of the class I MHC-HLA2 complex bound to the HIV reverse transcriptase epitope, ILKEPVHGV, and in the absence of the epitope. Each simulation was extended for 5ns. We studied the processes of water penetration in the interface between MHC and peptide, and identified 14 water molecules that stay bound for periods longer than 1ns in regions previously identified by crystallography. These water molecules in the interface perform definite "tasks" contributing to the binding energy: hydrogen bond bridges between MHC and peptide and filling empty spaces in the groove which enhance affinity without contributing to epitope specificity. We calculate the binding energy for interfacial water molecules and find that there is an overall gain in free energy resulting from the formation of water clusters at the epitope-MHC interface. Water molecules serving the task of filling empty spaces bind at the interface with a net gain in entropy, relative to their entropy in bulk. We conclude that water molecules at the interface play the role of active mediators in the MHC-peptide interaction, and might be responsible for the large binding affinity of the MHC complex to a large number of epitope sequences.

Biodiversity of small molecules--a new perspective in screening set selection.

How is the 'diversity' of a compound set defined and how is the most appropriate compound subset identified for assay when screening the entire HTS deck is not an option? A common approach has so far been to cover as much of the chemical space as possible by screening a chemically diverse set of compounds. We show that, rather than chemical diversity, the biologic diversity of a compound library is an essential requirement for hit identification. We describe a simple and efficient approach for the design of a HTS library based on compound-target diversity. Biodiverse compound subsets outperform chemically diverse libraries regarding hit rate and the total number of unique chemical scaffolds present among hits. Specifically, by screening ~19% of a HTS collection, we expect to discover ~50-80% of all desired bioactive compounds.

Rethinking molecular similarity: comparing compounds on the basis of biological activity.

Since the advent of high-throughput screening (HTS), there has been an urgent need for methods that facilitate the interrogation of large-scale chemical biology data to build a mode of action (MoA) hypothesis. This can be done either prior to the HTS by subset design of compounds with known MoA or post HTS by data annotation and mining. To enable this process, we developed a tool that compares compounds solely on the basis of their bioactivity: the chemical biological descriptor "high-throughput screening fingerprint" (HTS-FP). In the current embodiment, data are aggregated from 195 biochemical and cell-based assays developed at Novartis and can be used to identify bioactivity relationships among the in-house collection comprising ~1.5 million compounds. We demonstrate the value of the HTS-FP for virtual screening and in particular scaffold hopping. HTS-FP outperforms state of the art methods in several aspects, retrieving bioactive compounds with remarkable chemical dissimilarity to a probe structure. We also apply HTS-FP for the design of screening subsets in HTS. Using retrospective data, we show that a biodiverse selection of plates performs significantly better than a chemically diverse selection of plates, both in terms of number of hits and diversity of chemotypes retrieved. This is also true in the case of hit expansion predictions using HTS-FP similarity. Sets of compounds clustered with HTS-FP are biologically meaningful, in the sense that these clusters enrich for genes and gene ontology (GO) terms, showing that compounds that are bioactively similar also tend to target proteins that operate together in the cell. HTS-FP are valuable not only because of their predictive power but mainly because they relate compounds solely on the basis of bioactivity, harnessing the accumulated knowledge of a high-throughput screening facility toward the understanding of how compounds interact with the proteome.