Search

Molecular search is important in chemistry, biology, and informatics for identifying molecular structures within large data sets, improving knowledge discovery and innovation, and making chemical data FAIR (findable, accessible, interoperable, reusable). Search algorithms for polymers are significantly less developed than those for small molecules because polymer search relies on searching by polymer name, which can be challenging because polymer naming is overly broad (i.e., polyethylene), complicated for complex chemical structures, and often does not correspond to official IUPAC conventions. Chemical structure search in polymers is limited to substructures, such as monomers, without awareness of connectivity or topology. This work introduces a novel query language and graph traversal search algorithm for polymers that provides the first search method able to fully capture all of the chemical structures present in polymers. The BigSMARTS query language, an extension of the small-molecule SMARTS language, allows users to write queries that localize monomer and functional group searches to different parts of the polymer, like the middle block of a triblock, the side chain of a graft, and the backbone of a repeat unit. The substructure search algorithm is based on the traversal of graph representations of the generating functions for the stochastic graphs of polymers. Operationally, the algorithm first identifies cycles representing the monomers and then the end groups and finally performs a depth-first search to match entire subgraphs. To validate the algorithm, hundreds of queries were searched against hundreds of target chemistries and topologies from the literature, with approximately 440,000 query-target pairs. This tool provides a detailed algorithm that can be implemented in search engines to provide search results with full matching of the monomer connectivity and polymer topology.

Glycoproteomics is a powerful yet analytically challenging research tool. Software packages aiding the interpretation of complex glycopeptide tandem mass spectra have appeared, but their relative performance remains untested. Conducted through the HUPO Human Glycoproteomics Initiative, this community study, comprising both developers and users of glycoproteomics software, evaluates solutions for system-wide glycopeptide analysis. The same mass spectrometrybased glycoproteomics datasets from human serum were shared with participants and the relative team performance for N- and O-glycopeptide data analysis was comprehensively established by orthogonal performance tests. Although the results were variable, several high-performance glycoproteomics informatics strategies were identified. Deep analysis of the data revealed key performance-associated search parameters and led to recommendations for improved 'high-coverage' and 'high-accuracy' glycoproteomics search solutions. This study concludes that diverse software packages for comprehensive glycopeptide data analysis exist, points to several high-performance search strategies and specifies key variables that will guide future software developments and assist informatics decision-making in glycoproteomics., (© 2021. The Author(s).)

Minichromosome maintenance complex component 7 (MCM7) belongs to the minichromosome maintenance family that is important for the initiation of eukaryotic DNA replication. Overexpression of the MCM7 protein is relative to cellular proliferation and responsible for aggressive malignancy in various cancers. Mechanistically, inhibition of MCM7 significantly reduces the cellular proliferation associated with cancer. To date, no effective small molecular candidate has been identified that can block the progression of cancer induced by the MCM7 protein. Therefore, the study has been designed to identify small molecular-like natural drug candidates against aggressive malignancy associated with various cancers by targeting MCM7 protein. To identify potential compounds against the targeted protein a comprehensive in silico drug design including molecular docking, ADME (Absorption, Distribution, Metabolism and Excretion), toxicity, and molecular dynamics (MD) simulation approaches has been applied. Seventy phytochemicals isolated from the neem tree (Azadiractha indica) were retrieved and screened against MCM7 protein by using the molecular docking simulation method, where the top four compounds have been chosen for further evaluation based on their binding affinities. Analysis of ADME and toxicity properties reveals the efficacy and safety of the selected four compounds. To validate the stability of the protein-ligand complex structure MD simulations approach has also been performed to the protein-ligand complex structure, which confirmed the stability of the selected three compounds including CAS ID:105377-74-0, CID:12308716 and CID:10505484 to the binding site of the protein. In the study, a comprehensive data screening process has performed based on the docking, ADMET properties, and MD simulation approaches, which found a good value of the selected four compounds against the targeted MCM7 protein and indicates as a promising and effective human anticancer agent., (© The Author(s) 2021. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com.)

Several diagnostic tools have been developed for clinical and epidemiological assays. RT-PCR and antigen detection tests are more useful for diagnosis of acute disease, while antibody tests allow the estimation of exposure in the population. Currently, there is an urgent need for the development of diagnostic tests for COVID-19 that can be used for large-scale epidemiological sampling. Through a comprehensive strategy, potential 16 mer antigenic peptides suited for antibody-based SARS-CoV-2 diagnosis were identified. A systematic scan of the three structural proteins (S,N and M) and the non-structural proteins (ORFs) present in the SARS-CoV-2 virus was conducted through the combination of immunoinformatic methods, peptide SPOT synthesis and an immunoassay with cellulose-bound peptides (Pepscan). The Pepscan filter paper sheets with synthetic peptides were tested against pools of sera of COVID-19 patients. Antibody recognition showed a strong signal for peptides corresponding to the S, N and M proteins of SARS-CoV-2 virus, but not for the ORFs proteins. The peptides exhibiting higher signal intensity were found in the C-terminal region of the N protein. Several peptides of this region showed strong recognition with all three immunoglobulins in the pools of sera. The differential reactivity observed between the different immunoglobulin isotypes (IgA, IgM and IgG) within different regions of the S and N proteins, can be advantageous for ensuring accurate diagnosis of all infected patients, with different times of exposure to infection. Few peptides of the M protein showed antibody recognition and no recognition was observed for peptides of the ORFs proteins., (Copyright © 2021 Elsevier B.V. All rights reserved.)

The Drosophila type II neuroblast lineages present an attractive model to investigate the neurogenesis and differentiation process as they adapt to a process similar to that in the human outer subventricular zone. We perform targeted single-cell mRNA sequencing in third instar larval brains to study this process of the type II NB lineage. Combining prior knowledge, in silico analyses, and in situ validation, our multi-informatic investigation describes the molecular landscape from a single developmental snapshot. 17 markers are identified to differentiate distinct maturation stages. 30 markers are identified to specify the stem cell origin and/or cell division numbers of INPs, and at least 12 neuronal subtypes are identified. To foster future discoveries, we provide annotated tables of pairwise gene-gene correlation in single cells and MiCV, a web tool for interactively analyzing scRNA-seq datasets. Taken together, these resources advance our understanding of the neural differentiation process at the molecular level., Competing Interests: Declaration of interests N.S.M. is the founder of MiOmics Inc. (MI, USA)., (Copyright © 2021 The Authors. Published by Elsevier Inc. All rights reserved.)

Weighted multiscale support vector regression combined with ultraviolet-visible (UV-Vis) spectra for quantitative analysis of edible blend oil is proposed. In the approach, UV-Vis spectra of the training set are decomposed into a certain number of intrinsic mode functions (IMFs) and a residue by empirical mode decomposition (EMD) at first. Then support vector regression (SVR) sub-models are built on each IMF and residue. For prediction set, the spectra are decomposed as done on the training set and the final predictions are obtained by integrating SVR sub-model predictions by weighted average. The weight of the sub-model is the reciprocal of the fourth power of the root mean square error of cross-validation (RMSECV). For predicting peanut oil in binary blend oil and sesame oil in ternary blend oil, the proposed method has superiority in root mean square error of prediction (RMSEP) and correlation coefficient (R) compared with SVR and partial least squares (PLS)., (Copyright © 2020 Elsevier Ltd. All rights reserved.)

Introduction: Transforming administrative health care data into meaningful metrics has been critical to the implementation of the Department of Defense's Primary Care Behavioral Health (PCBH) program., Methods: Data from clinical encounters with PCBH providers are used to develop metrics of program performance collaboratively. Metrics focus on describing the PCBH program and patients, provider fidelity to the model, and provider performance. These metrics form two key deliverables: a monitoring dashboard for program managers and a training dashboard for expert trainers conducting site visits., Results: Behavioral health consultants (BHCs) conducted nearly 200,000 encounters with more than 100,000 unique patients in fiscal year 2019 at more than 170 locations in 6 countries and 37 states. Administrative data derived from these encounters were used to create a variety of metrics that describe practice and performance at both the provider and program levels. These metrics are delivered through a variety of analytic products to stakeholders who use that information to make data-driven decisions about program direction and provider training., Discussion: We discuss examples of program management decisions and expert trainer actions based on these dashboards, highlighting the benefits of continued collaboration between analysts and program managers. Specifically, excerpts from several dashboards illustrate how penetration and productivity metrics yield specific, tailored action plans to improve care delivery and provider performance. (PsycInfo Database Record (c) 2021 APA, all rights reserved).

While over 150 distinct types of chemical modifications are known to occur on various cellular RNAs and can be dynamically controlled, the function of most of these modifications remains poorly defined. Collectively, these RNA modifications have been recently termed the "epitranscriptome". Identification and annotation of individual RNA modifications throughout the transcriptome are key for studying the role of the epitranscriptome in the regulation of gene expression and for elucidating the functional relevance of particular RNA modifications in diverse physiological and disease processes. In this protocol, we demonstrate how to identify and annotate RNA modifications based on the informatic analysis of methylated RNA immunoprecipitation and sequencing (MeRIP-seq) data, using RNAmod, a convenient one-stop online interactive platform for the annotation, analysis, and visualization of mRNA modifications.

Computer and information systems can improve occupational respiratory disease prevention and surveillance by providing efficient resources for patients, workers, clinicians, and public health practitioners. Advances include interlinking electronic health records, autocoding surveillance data, clinical decision support systems, and social media applications for acquiring and disseminating information. Obstacles to advances include inflexible hierarchical coding schemes, inadequate occupational health electronic health record systems, and inadequate public focus on occupational respiratory disease. Potentially transformative approaches include machine learning, natural language processing, and improved ontologies., Competing Interests: Disclosure The authors have no conflict of interest, nor specific funding for this work., (Copyright © 2020 Elsevier Inc. All rights reserved.)

The utility of an autoencoder (AE) as a feature extraction tool for near-infrared (NIR) spectroscopy-based discrimination analysis has been explored and the discrimination of the geographic origins of 8 different agricultural products has been performed as the case study. The sample spectral features were broad and insufficient for component distinction due to considerable overlap of individual bands, so AE enabling of extracting the sample-descriptive features in the spectra would help to improve discrimination accuracy. For comparison, four different inputs of AE-extracted features, raw NIR spectra, principal component (PC) scores, and features extracted using locally linear embedding were employed for sample discrimination using support vector machine. The use of AE-extracted feature improved the accuracy in the discrimination of samples in all 8 products. The improvement was more substantial when the sample spectral features were indistinct. It demonstrates that AE is expandable for vibrational spectroscopic discriminant analysis of other samples with complex composition., Competing Interests: Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2020 Elsevier Ltd. All rights reserved.)

Phenotyping mouse model systems of human disease has proven to be a difficult task, with frequent poor inter- and intra-laboratory replicability, particularly in behavioral domains such as social and cognitive function. However, establishing robust animal model systems with strong construct validity is of fundamental importance as they are central tools for understanding disease pathophysiology and developing therapeutics. To complete our studies of mouse model systems relevant to autism spectrum disorder (ASD), we present a replication of the main findings from our two published studies of five genetic mouse model systems of ASD. To assess the intra-laboratory robustness of previous results, we chose the two model systems that showed the greatest phenotypic differences, the Shank3/F and Cntnap2, and repeated assessments of general health, activity and social behavior. We additionally explored all five model systems in the same framework, comparing all results obtained in this three-yearlong effort using informatics techniques to assess commonalities and differences. Our results showed high intra-laboratory replicability of results, even for those with effect sizes that were not particularly large, suggesting that discrepancies in the literature may be dependent on subtle but pivotal differences in testing conditions, housing enrichment, or background strains and less so on the variability of the behavioral phenotypes. The overall informatics analysis suggests that in our behavioral assays we can separate the set of tested mouse model system into two main classes that in some aspects lie on opposite ends of the behavioral spectrum, supporting the view that autism is not a unitary concept., (© 2020 The Authors. Genes, Brain and Behavior published by International Behavioural and Neural Genetics Society and John Wiley & Sons Ltd.)

Researchers must collaborate globally to rapidly respond to the COVID-19 pandemic. In Europe, the General Data Protection Regulation (GDPR) regulates the processing of personal data, including health data of value to researchers. Even during a pandemic, research still requires a legal basis for the processing of sensitive data, additional justification for its processing, and a basis for any transfer of data outside Europe. The GDPR does provide legal grounds and derogations that can support research addressing a pandemic, if the data processing activities are proportionate to the aim pursued and accompanied by suitable safeguards. During a pandemic, a public interest basis may be more promising for research than a consent basis, given the high standards set out in the GDPR. However, the GDPR leaves many aspects of the public interest basis to be determined by individual Member States, which have not fully or uniformly made use of all options. The consequence is an inconsistent legal patchwork that displays insufficient clarity and impedes joint approaches. The COVID-19 experience provides lessons for national legislatures. Responsiveness to pandemics requires clear and harmonized laws that consider the related practical challenges and support collaborative global research in the public interest., (©Regina Becker, Adrian Thorogood, Johan Ordish, Michael J.S. Beauvais. Originally published in the Journal of Medical Internet Research (http://www.jmir.org), 27.08.2020.)

Despite the widespread use of the "Informatics for Integrating Biology and the Bedside" (i2b2) platform, there are substantial challenges for loading electronic health records (EHR) into i2b2 and for querying i2b2. We have previously presented a simplified framework for semantic abstraction of EHR records into i2b2. Building on our previous work, we have created a proof-of-concept implementation of cloud services on an i2b2 data store for cohort identification. Specifically, we have implemented a graphical user interface (GUI) that declares the key components for data import, transformation, and query of EHR data. The GUI integrates with Azure cloud services to create data pipelines for importing EHR data into i2b2, creation of derived facts, and querying for generating Sankey-like flow diagrams that characterize the patient cohorts. We have evaluated the implementation using the real-world MIMIC-III dataset. We discuss the key features of this implementation and direction for future work, which will advance the efforts of the research community for patient cohort identification., Competing Interests: The authors declare that they have no conflicts of interest., (Copyright © 2020 Kavishwar B. Wagholikar et al.)

Background: The language gap between health consumers and health professionals has been long recognized as the main hindrance to effective health information comprehension. Although providing health information access in consumer health language (CHL) is widely accepted as the solution to the problem, health consumers are found to have varying health language preferences and proficiencies. To simplify health documents for heterogeneous consumer groups, it is important to quantify how CHLs are different in terms of complexity among various consumer groups., Objective: This study aimed to propose an informatics framework (consumer health language complexity [CHELC]) to assess the complexity differences of CHL using syntax-level, text-level, term-level, and semantic-level complexity metrics. Specifically, we identified 8 language complexity metrics validated in previous literature and combined them into a 4-faceted framework. Through a rank-based algorithm, we developed unifying scores (CHELC scores [CHELCS]) to quantify syntax-level, text-level, term-level, semantic-level, and overall CHL complexity. We applied CHELCS to compare posts of each individual on online health forums designed for (1) the general public, (2) deaf and hearing-impaired people, and (3) people with autism spectrum disorder (ASD)., Methods: We examined posts with more than 4 sentences of each user from 3 health forums to understand CHL complexity differences among these groups: 12,560 posts from 3756 users in Yahoo! Answers, 25,545 posts from 1623 users in AllDeaf, and 26,484 posts from 2751 users in Wrong Planet. We calculated CHELCS for each user and compared the scores of 3 user groups (ie, deaf and hearing-impaired people, people with ASD, and the public) through 2-sample Kolmogorov-Smirnov tests and analysis of covariance tests., Results: The results suggest that users in the public forum used more complex CHL, particularly more diverse semantics and more complex health terms compared with users in the ASD and deaf and hearing-impaired user forums. However, between the latter 2 groups, people with ASD used more complex words, and deaf and hearing-impaired users used more complex syntax., Conclusions: Our results show that the users in 3 online forums had significantly different CHL complexities in different facets. The proposed framework and detailed measurements help to quantify these CHL complexity differences comprehensively. The results emphasize the importance of tailoring health-related content for different consumer groups with varying CHL complexities., (©Biyang Yu, Zhe He, Aiwen Xing, Mia Liza A Lustria. Originally published in the Journal of Medical Internet Research (http://www.jmir.org), 21.05.2020.)

Concerns about gender bias in word embedding models have captured substantial attention in the algorithmic bias research literature. Other bias types however have received lesser amounts of scrutiny. This work describes a large-scale analysis of sentiment associations in popular word embedding models along the lines of gender and ethnicity but also along the less frequently studied dimensions of socioeconomic status, age, physical appearance, sexual orientation, religious sentiment and political leanings. Consistent with previous scholarly literature, this work has found systemic bias against given names popular among African-Americans in most embedding models examined. Gender bias in embedding models however appears to be multifaceted and often reversed in polarity to what has been regularly reported. Interestingly, using the common operationalization of the term bias in the fairness literature, novel types of so far unreported bias types in word embedding models have also been identified. Specifically, the popular embedding models analyzed here display negative biases against middle and working-class socioeconomic status, male children, senior citizens, plain physical appearance and intellectual phenomena such as Islamic religious faith, non-religiosity and conservative political orientation. Reasons for the paradoxical underreporting of these bias types in the relevant literature are probably manifold but widely held blind spots when searching for algorithmic bias and a lack of widespread technical jargon to unambiguously describe a variety of algorithmic associations could conceivably be playing a role. The causal origins for the multiplicity of loaded associations attached to distinct demographic groups within embedding models are often unclear but the heterogeneity of said associations and their potential multifactorial roots raises doubts about the validity of grouping them all under the umbrella term bias. Richer and more fine-grained terminology as well as a more comprehensive exploration of the bias landscape could help the fairness epistemic community to characterize and neutralize algorithmic discrimination more efficiently., Competing Interests: The authors have declared that no competing interests exist.

Apathy is a common neuropsychiatric syndrome observed across many neurocognitive and psychiatric disorders. Although there are currently no definitive standard therapies for the treatment of apathy, nonpharmacological treatment (NPT) is often considered to be at the forefront of clinical management. However, guidelines on how to select, prescribe, and administer NPT in clinical practice are lacking. Furthermore, although new Information and Communication Technologies (ICT) are beginning to be employed in NPT, their role is still unclear. The objective of the present work is to provide recommendations for the use of NPT for apathy, and to discuss the role of ICT in this domain, based on opinions gathered from experts in the field. The expert panel included 20 researchers and healthcare professionals working on brain disorders and apathy. Following a standard Delphi methodology, experts answered questions via several rounds of web-surveys, and then discussed the results in a plenary meeting. The experts suggested that NPT are useful to consider as therapy for people presenting with different neurocognitive and psychiatric diseases at all stages, with evidence of apathy across domains. The presence of a therapist and/or a caregiver is important in delivering NPT effectively, but parts of the treatment may be performed by the patient alone. NPT can be delivered both in clinical settings and at home. However, while remote treatment delivery may be cost and time-effective, it should be considered with caution, and tailored based on the patient's cognitive and physical profile and living conditions., (Copyright © 2019 American Association for Geriatric Psychiatry. Published by Elsevier Inc. All rights reserved.)

The dawn of the 21st century has brought with it a surge of research related to computer-guided approaches to catalyst design. In the past two decades, chemoinformatics, the application of informatics to solve problems in chemistry, has increasingly influenced prediction of activity and mechanistic investigations of organic reactions. The advent of advanced statistical and machine learning methods, as well as dramatic increases in computational speed and memory, has contributed to this emerging field of study. This review summarizes strategies to employ quantitative structure-selectivity relationships (QSSR) in asymmetric catalytic reactions. The coverage is structured by initially introducing the basic features of these methods. Subsequent topics are discussed according to increasing complexity of molecular representations. As the most applied subfield of QSSR in enantioselective catalysis, the application of local parametrization approaches and linear free energy relationships (LFERs) along with multivariate modeling techniques is described first. This section is followed by a description of global parametrization methods, the first of which is continuous chirality measures (CCM) because it is a single parameter derived from the global structure of a molecule. Chirality codes, global, multivariate descriptors, are then introduced followed by molecular interaction fields (MIFs), a global descriptor class that typically has the highest dimensionality. To highlight the current reach of QSSR in enantioselective transformations, a comprehensive collection of examples is presented. When combined with traditional experimental approaches, chemoinformatics holds great promise to predict new catalyst structures, rationalize mechanistic behavior, and profoundly change the way chemists discover and optimize reactions.

Increased performance demands have interacted with suboptimal use of technology and contributed to burnout among radiologists. Although the problem of radiologist burnout has been well documented, there is a gap in the literature in terms of how technology can be better utilized to lessen the problem. Informatics-based modifications to existing technology hold the potential to reduce the amount of time radiologists spend on noninterpretive tasks, decrease interruptions, facilitate connections with colleagues, and improve patient care. Examples of successful modifications to technology are presented and discussed in relation to how they contribute to improving workplace engagement among radiologists., (Copyright © 2019 Elsevier Inc. All rights reserved.)

This study represents the first attempt to combine mid infrared (MIR) spectroscopy and multivariate data processing for prediction of alcohol degree, sugars content and total acidity in straw wine. 302 Italian samples, representing different vintages, production regions and grape varieties, were analysed using FT-MIR spectroscopy and reference methods. New regression functions based on a combination of Orthogonal Signal Correction and Partial Least Squares regression are proposed for prediction of quality parameters: this approach allows overcoming the issue of matrix complexity, reducing spectral interferences and enhancing the information embodied in fingerprinting data. The models proposed are characterised by an excellent reliability, with low error in prediction (alcohol: 0.28%; sugars: 9.9 g/L; acidity: 0.29 g/L) comparable both to reference methods and table wine models. Results demonstrate that vibrational spectroscopy, combined with a proper multivariate data strategy, represents a suitable strategy for the quick and non-destructive assessment of quality parameters of straw wine., (Copyright © 2019 Elsevier Ltd. All rights reserved.)

Eukaryotic algae are an extremely diverse category of photosynthetic organisms and some species produce highly potent bioactive compounds poisonous to humans or other animals, most notably observed during harmful algal blooms. These natural products include some of the most poisonous small molecules known and unique cyclic polyethers. However, the diversity and complexity of algal genomes means that sequencing-based research has lagged behind research into more readily sequenced microbes, such as bacteria and fungi. Applying informatics techniques to the algal genomes that are now available reveals new natural product biosynthetic pathways, with different groups of algae containing different types of pathways. There is some evidence for gene clusters and the biosynthetic logic of polyketides enables some prediction of these final products. For other pathways, it is much more challenging to predict the products and there may be many gene clusters that are not identified with the automated tools. These results suggest that there is a great diversity of biosynthetic capacity for natural products encoded in the genomes of algae and suggest areas for future research focus., Competing Interests: The author declares no conflict of interest

Animals are used to study the pathogenesis of various human diseases, but typically as animal models with induced disease. However, companion animals develop disease spontaneously in a way that mirrors disease development in humans. The purpose of this study is to develop a semantic and domain-specific method to enable construction of a data repository from a veterinary hospital that would be useful for future studies. We developed a two-phase method that combines semantic and domain-specific approaches to construct a canine data repository of clinical data collected during routine care at the Matthew J Ryan Veterinary Hospital of the University of Pennsylvania (PennVet). Our framework consists of two phases: (1) a semantic data-cleaning phase and (2) a domain-specific data-cleaning phase. We validated our data repository using a gold standard of known breed predispositions for certain diseases (i.e., mitral valve disease, atrial fibrillation and osteosarcoma). Our two-phase method allowed us to maximize data retention (99.8% of data retained), while ensuring the quality of our result. Our final population contained 84,405 dogs treated between 2000 and 2017 from 194 distinct dog breeds. We observed the expected breed associations with mitral valve disease, atrial fibrillation, and osteosarcoma (P < 0.05) after adjusting for multiple comparisons. Precision ranged from 60.0 to 83.3 for the three diseases (avg. 74.2) and recall ranged from 31.6 to 83.3 (avg. 53.3). Our study describes a two-phase method to construct a clinical data repository using canine data obtained during routine clinical care at a veterinary hospital.

Accurate prediction of molecular properties is important for new compound design, which is a crucial step in drug discovery. In this paper, molecular graph data is utilized for property prediction based on graph convolution neural networks. In addition, a convolution spatial graph embedding layer (C-SGEL) is introduced to retain the spatial connection information on molecules. And, multiple C-SGELs are stacked to construct a convolution spatial graph embedding network (C-SGEN) for end-to-end representation learning. In order to enhance the robustness of the network, molecular fingerprints are also combined with C-SGEN to build a composite model for predicting molecular properties. Our comparative experiments have shown that our method is accurate and achieves the best results on some open benchmark datasets.

The health promoting effects of extra virgin olive oil (EVOO) relate to its unique repertoire of phenolic compounds. Here, we used a chemoinformatics approach to computationally identify endogenous ligands and assign putative biomolecular targets to oleacein, one of the most abundant secoiridoids in EVOO. Using a structure-based virtual profiling software tool and reference databases containing more than 9000 binding sites protein cavities, we identified 996 putative oleacein targets involving more than 700 proteins. We subsequently identified the high-level functions of oleacein in terms of biomolecular interactions, signaling pathways, and protein-protein interaction (PPI) networks. Delineation of the oleacein target landscape revealed that the most significant modules affected by oleacein were associated with metabolic processes (e.g., glucose and lipid metabolism) and chromatin-modifying enzymatic activities (i.e., histone post-translational modifications). We experimentally confirmed that, in a low-micromolar physiological range (<20 μmol/l), oleacein was capable of inhibiting the catalytic activities of predicted metabolic and epigenetic targets including nicotinamide N-methyltransferase, ATP-citrate lyase, lysine-specific demethylase 6A, and N-methyltransferase 4. Our computational de-orphanization of oleacein provides new mechanisms through which EVOO biophenols might operate as chemical prototypes capable of modulating the biologic machinery of healthy aging., (Copyright © 2019 Elsevier Ltd. All rights reserved.)

It is the position of the Academy of Nutrition and Dietetics that nutrition informatics is a rapidly evolving area of practice for registered dietitian nutritionists and nutrition and dietetic technicians, registered; and that the knowledge and skills inherent to nutrition informatics permeate all areas of the dietetics profession. Further, nutrition and dietetics practitioners must continually learn and update their informatics knowledge and skills to remain at the forefront of nutrition practice. Nutrition informatics is the intersection of information, nutrition, and technology. However, informatics is not just using technology to do work. The essence of nutrition informatics is to manage nutrition data in combination with standards, processes, and technology to improve knowledge and practice that ultimately lead to improved quality of health care and work efficiency. Registered dietitian nutritionists and nutrition and dietetic technicians, registered, are already experts in using evidence to practice in all areas of nutrition and dietetics. To remain at the forefront of technological innovation, the profession must actively participate in the development of standards, processes, and technologies for providing nutrition care., (Copyright © 2019 Academy of Nutrition and Dietetics. Published by Elsevier Inc. All rights reserved.)

The reuse of business processes (BPs) requires similarities between them to be suitably identified. Various approaches have been introduced to address this problem, but many of them feature a high computational cost and a low level of automation. This paper presents a clustering algorithm that groups business processes retrieved from a multimodal search system (based on textual and structural information). The algorithm is based on Incremental Covering Arrays (ICAs) with different alphabets to determine the possible number of groups to be created for each row of the ICA. The proposed algorithm also incorporates Balanced Bayesian Information Criterion to determine the optimal number of groups and the best solution for each query. Experimental evaluation shows that the use of ICAs with strength four (4) and different alphabets reduces the number of solutions needed to be evaluated and optimizes the number of clusters. The proposed algorithm outperforms other algorithms in various measures (precision, recall, and F-measure) by between 12% and 88%. Friedman and Wilcoxon non-parametric tests gave a 90-95% significance level to the obtained results. Better options of repository search for BPs help companies to reuse them. By thus reusing BPs, managers and analysts can more easily get to know the evolution and trajectory of the company processes, a situation that could be expected to lead to improved managerial and commercial decision making., Competing Interests: The authors have declared that no competing interests exist.

Background: Representation learning provides new and powerful graph analytical approaches and tools for the highly valued data science challenge of mining knowledge graphs. Since previous graph analytical methods have mostly focused on homogeneous graphs, an important current challenge is extending this methodology for richly heterogeneous graphs and knowledge domains. The biomedical sciences are such a domain, reflecting the complexity of biology, with entities such as genes, proteins, drugs, diseases, and phenotypes, and relationships such as gene co-expression, biochemical regulation, and biomolecular inhibition or activation. Therefore, the semantics of edges and nodes are critical for representation learning and knowledge discovery in real world biomedical problems., Results: In this paper, we propose the edge2vec model, which represents graphs considering edge semantics. An edge-type transition matrix is trained by an Expectation-Maximization approach, and a stochastic gradient descent model is employed to learn node embedding on a heterogeneous graph via the trained transition matrix. edge2vec is validated on three biomedical domain tasks: biomedical entity classification, compound-gene bioactivity prediction, and biomedical information retrieval. Results show that by considering edge-types into node embedding learning in heterogeneous graphs, edge2vec significantly outperforms state-of-the-art models on all three tasks., Conclusions: We propose this method for its added value relative to existing graph analytical methodology, and in the real world context of biomedical knowledge discovery applicability.

Pharmacophore models in general use a variety of features for distinct chemical characteristics, such as hydrogen-bond properties, lipohilicity, and ionizability. Usually, features have to match onto their identical type. To clarify if this stringent one-to-one assignment is justified, we investigated a set of 581 unique ligands from the BindingDB with known orientation inside the respective binding pockets and conducted a statistical analysis of the likelihood of observed exchanges in between the pharmacophore features, respectively their degree of conservation. To find out if certain features are obsolete, we derived a ranking to determine the most relevant ones. We found that the most conserved one-to-one feature is the negative ionizable (acids), followed by hydrogen-bond donor, positive ionizable (basic nitrogens), hydrogen-bond acceptor, aromatic, nonaromatic π-systems, and other lipophilic characteristics. The most likely exchanges were found between carboxylate groups and hydrogen-bond acceptors and likewise between basic nitrogens and hydrogen-bond donors, which reflects the characteristics of Lewis acids and bases. Exchanges between hydrogen-bond donors and hydrogen-bond acceptors are hardly more likely than by chance. The kind of target (e.g., kinase, phosphatase, protease, phosphodiesterase, nuclear receptor, metal-containing, or transmembrane protein) did not show substantial influence on the degree of conservation. The relevance of the actual pharmacophore features was found to be strongly dependent on the applied ranking scheme. Mutual information ranks all hydrophobic features as least important, whereas the aromatic feature is put into second place by using a geometric series. Both ranking schemes see the negative ionizable feature of higher significance than the positively ionizable feature.

Binding prediction between targets and drug-like compounds through deep neural networks has generated promising results in recent years, outperforming traditional machine learning-based methods. However, the generalization capability of these classification models is still an issue to be addressed. In this work, we explored how different cross-validation strategies applied to data from different molecular databases affect to the performance of binding prediction proteochemometrics models. These strategies are (1) random splitting, (2) splitting based on K-means clustering (both of actives and inactives), (3) splitting based on source database, and (4) splitting based both in the clustering and in the source database. These schemas are applied to a deep learning proteochemometrics model and to a simple logistic regression model to be used as baseline. Additionally, two different ways of describing molecules in the model are tested: (1) by their SMILES and (2) by three fingerprints. The classification performance of our deep learning-based proteochemometrics model is comparable to the state of the art. Our results show that the lack of generalization of these models is due to a bias in public molecular databases and that a restrictive cross-validation schema based on compound clustering leads to worse but more robust and credible results. Our results also show better performance when representing molecules by their fingerprints.

Multitask prediction of bioactivities is often faced with challenges relating to the sparsity of data and imbalance between different labels. We propose class conditional (Mondrian) conformal predictors using underlying Macau models as a novel approach for large scale bioactivity prediction. This approach handles both high degrees of missing data and label imbalances while still producing high quality predictive models. When applied to ten assay end points from PubChem, the models generated valid models with an efficiency of 74.0-80.1% at the 80% confidence level with similar performance both for the minority and majority class. Also when deleting progressively larger portions of the available data (0-80%) the performance of the models remained robust with only minor deterioration (reduction in efficiency between 5 and 10%). Compared to using Macau without conformal prediction the method presented here significantly improves the performance on imbalanced data sets.

In practice, most audio files such as MP3 and AAC are stored and transmitted in the form of compressed files, which can serve as the cover in audio steganography. Currently, the prevailing audio steganography methods are not ideal because of the drawbacks. Some are characterized by low capacity while others are irreversible. In this paper, we propose a method to embed secret messages in MP3 encoding. Our strategy is to hide the information by Huffman table transformation. We extract secret information by analyzing side information. Experimental results show that our method can greatly improve the steganographic capacity with low distortion and high security. Meanwhile, it featured with higher decoding rate and reversible over some state-of-the-art methods.

Despite the increasing volume of available data, the proportion of experimentally measured data remains small compared to the virtual chemical space of possible chemical structures. Therefore, there is a strong interest in simultaneously predicting different ADMET and biological properties of molecules, which are frequently strongly correlated with one another. Such joint data analyses can increase the accuracy of models by exploiting their common representation and identifying common features between individual properties. In this work we review the recent developments in multi-learning approaches as well as cover the freely available tools and packages that can be used to perform such studies., (© 2018 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA.)

Background: Primary Ovarian Insufficiency (POI), a major cause of infertility, affects about 1-3% of women under forty years of age. Although there is a growing list of causal genetic alterations, POI remains mostly idiopathic., Methods: We performed exome sequencing (WES) of two sisters affected with POI, one unaffected sister and their mother from a consanguineous family. We assessed the impact of the identified MEIOB variant with a minigene assay and by sequencing illegitimate transcripts from the proband's leukocytes. We studied its functional impact on the interaction between MEIOB with its partner SPATA22 and their localization to DNA double-strand breaks (DSB)., Findings: We identified a homozygous variant in the last base of exon 12 of MEIOB, which encodes a factor essential for meiotic recombination. This variant was predicted to strongly affect MEIOB pre-mRNA splicing. Consistently, a minigene assay showed that the variant induced exon 12 skipping, which was confirmed in vivo in the proband's leukocytes. Aberrant splicing leads to the production of a C-terminally truncated protein that cannot interact with SPATA22, abolishing their recruitment to DSBs., Interpretation: This truncating MEIOB variant is expected to provoke meiotic defects and a depleted follicular stock, as in Meiob -/- mice. This is the first molecular defect reported in a meiosis-specific single-stranded DNA-binding protein (SSB) responsible for POI. We hypothesise that alterations in other SSB proteins could explain cases of syndromic or isolated ovarian insufficiency. FUND: Université Paris Diderot, Fondation pour la Recherche Médicale, Fondation ARC contre le cancer, Commissariat à l'Energie Atomique and Institut Universitaire de France., (Copyright © 2019 The Authors. Published by Elsevier B.V. All rights reserved.)

A metaproteomic analysis was conducted on the fecal microbiome of eight infants to characterize global protein and pathway expression. Although mass spectrometry-based proteomics is now a routine tool, analysis of the microbiome presents specific technical challenges, including the complexity and dynamic range of member taxa, the need for well-annotated metagenomic databases, and high inter-protein sequence redundancy and similarity. In this study, an approach was developed for assessment of biological phenotype and metabolic status, as a functional complement to DNA sequence analysis. Fecal samples were prepared and analysed by tandem mass spectrometry and a homology-based meta-clustering strategy was used to combine peptides from multiple species into representative proteins. In total, 15,250 unique peptides were sequenced and assigned to 2154 metaclusters, which were then assigned to pathways and functional groups. Differences were noted in several pathways, consistent with the dominant genera observed in different subjects. Although this study was not powered to draw conclusions from the comparisons, the results obtained demonstrate the applicability of this approach and provide the methods needed for performing semi-quantitative comparisons of human fecal microbiome composition, physiology and metabolism, as well as a more detailed assessment of microbial composition in comparison to 16S rRNA gene sequencing.

Existing distortion functions in steganography which achieved high undetectability are designed for unprocessed natural image. Nowadays, a large number of images are filtered before transmitting for the sake of beautification. In this situation, existing distortion functions should be improved to fit the properties of these beautified images. This paper proposes a distortion function optimization method for steganography on beautified images. Given an unprocessed image, a popular image beautification software is employed to produce two similar beautified images. One of them is used for embedding and the other one is employed as reference. Guided by the reference, existing distortion functions are improved by distinguishing the embedding costs for ±1 embedding. After embedding, the stego image is closer to the reference, which results in a higher undetectability against steganalysis. Experimental results also proved the increasing of undetectability when examined by modern steganalytic tools.

Edge computing is an important tool for smart computing, which brings convenience to data processing as well as security problems. In particular, the security of data storage under edge computing has become an obstacle to its widespread use. To solve the problem, the mechanism combing blockchain with regeneration coding is proposed to improve the security and reliability of stored data under edge computing. Our contribution is as follows. 1) According to the three-tier edge computing architecture and data security storage requirements, we proposed hybrid storage architecture and model specifically adapted to edge computing. 2) Making full use of the data storage advantages of edge network devices and cloud storage servers, we build a global blockchain in the cloud service layer and local blockchain is built on the terminals of the Internet of things. Moreover, the regeneration coding is utilized to further improve the reliability of data storage in blockchains. 3) Our scheme provides a mechanism for periodically validating hash values of data to ensure the integrity of data stored in global blockchain.

We present DrugScore 2018 , a new version of the knowledge-based scoring function DrugScore, which builds upon the same formalism used to derive DrugScore but exploits a training data set of nearly 40 000 X-ray complex structures, a highly diverse and the, by far, largest data set ever used for such an endeavor. About 2.5 times as many pair potentials than before now have a data basis required to yield smooth potentials, and pair potentials could now be derived for eight more atom types, including interactions involving halogen atoms and metal ions highly relevant for medicinal chemistry. Probing for dependence on training data set size and quality, we show that DrugScore 2018 potentials are converged. We evaluated DrugScore 2018 in comprehensive scoring, ranking, docking, and screening tests on the CASF-2013 data set, allowing for a comparison with >30 other scoring functions. There, DrugScore 2018 showed similar or improved performance in all aspects when compared to either DrugScore, DrugScore CSD , or DSX and was, overall, the scoring function showing the most consistently good performance in scoring, ranking, and docking tests. Applying DrugScore 2018 as objective function in AutoDock3 in a large-scale docking trial, using 4056 protein-ligand complexes from PDBbind 2016, reproduced a docked pose to within 2 Å RMSD to the crystal structure in >75% of all dockings. These results are remarkable as the DrugScore 2018 potentials were derived from crystallographic information only, without any further adaptation using binding affinity or docking decoy data. DrugScore 2018 should thus be a competitive scoring and objective function for structure-based ligand design purposes.

Background: The digitization of society has an increasing impact on healthcare in general and, therefore, also on psychiatry.
AIM: To provide an overview of digital developments and their influence on the design of future professional psychiatric care.
METHOD: With the help of examples from literature, show how digitization will influence diagnostic procedures as well as psychiatric treatment.
RESULTS: Digitization will have a major impact on psychiatric diagnostics and treatment. For example, psychiatric diagnostics will be strengthened by continuous monitoring of behaviour with digital wearables and the collection of large amounts of personal data. How we deal with these new sources of information needs to be developed in everyday practice. Psychiatric treatments with E-health, online therapies, apps and virtual reality are being developed rapidly. There is increasing evidence concerning the efficacy of these treatments in a variety of patient groups.
CONCLUSION: The digital revolution in psychiatric health services has just begun. To maximise the benefits of digitization for psychiatry, it is necessary to connect technological possibilities with well-founded scientific knowledge, professional expertise, expectations and needs of patients, and clear legal instructions.

In medicinal chemistry, lead optimization is a critically important task and a highly empirical process, largely driven by chemical knowledge and intuition. Only very few approaches are available to guide and evaluate optimization efforts. It is often very difficult to understand when a compound series is exhausted and the generation of additional analogs unlikely to yield further progress toward potent and efficacious candidates. Rationalizing lead optimization remains an essentially unsolved problem. Herein, we introduce a new computational method to aid in evaluating whether sufficient numbers of analogs have been made and further progress is unlikely. The approach integrates the assessment of chemical saturation and structure-activity relationship progression of compound series. Easy-to-calculate scores characterize evolving analog series and identify candidates with high or low priority for further chemical exploration.

Cardiomyocytes derived from human pluripotent stem cells are a promising tool for disease modeling, drug compound testing, and cardiac toxicity screening. Bio-image segmentation is a prerequisite step in cardiomyocyte image analysis by digital holography (DH) in microscopic configuration and has provided satisfactory results. In this study, we quantified multiple cardiac cells from segmented 3-dimensional DH images at the single-cell level and measured multiple parameters describing the beating profile of each individual cell. The beating profile is extracted by monitoring dry-mass distribution during the mechanical contraction-relaxation activity caused by cardiac action potential. We present a robust two-step segmentation method for cardiomyocyte low-contrast image segmentation based on region and edge information. The segmented single-cell contains mostly the nucleus of the cell since it is the best part of the cardiac cell, which can be perfectly segmented. Clustering accuracy was assessed by a silhouette index evaluation for k-means clustering and the Dice similarity coefficient (DSC) of the final segmented image. 3D representation of single of cardiomyocytes. The cell contains mostly the nucleus section and a small area of cytoplasm., (© 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim.)

Murine herpesvirus-68 (MHV-68) productively infects mouse lungs, exhibiting a complex pathology characteristic of both acute viral infections and chronic respiratory diseases. We sought to discover proteins differentially expressed in bronchoalveolar lavage (BAL) from mice infected with MHV-68. Mice were infected intranasally with MHV-68. After nine days, as the lytic phase of infection resolved, differential BAL proteins were identified by two-dimensional (2D) electrophoresis and mass spectrometry. Of 23 unique proteins, acute phase proteins, vitamin A transport, and oxidative stress response factors Pdx6 and EC-SOD (Sod3) were enriched. Correspondingly, iNOS2 was induced in lung tissue by seven days post-infection. Oxidative stress was partly a direct result of MHV-68 infection, as reactive oxygen species (ROS) were induced in cultured murine NIH3T3 fibroblasts and human lung A549 cells infected with MHV-68. Finally, mice infected with a recombinant MHV-68 co-expressing inflammatory cytokine murine interleukin 6 (IL6) showed exacerbated oxidative stress and soluble type I collagen characteristic of tissue recovery. Thus, oxidative stress appears to be a salient feature of MHV-68 pathogenesis, in part caused by lytic replication of the virus and IL6. Proteins and small molecules in lung oxidative stress networks therefore may provide new therapeutic targets to ameliorate respiratory virus infections.

Purpose: Non-muscle-invasive bladder cancer (NMIBC) is a malignant disease characterized by high heterogeneity, which corresponds to dysregulated gene expression and alternative splicing (AS) profiles. Bioinformatics analyses of splicing factors potentially linked to bladder cancer progression identified the heterogeneous nuclear ribonucleoprotein I (i.e., PTBP1) as candidate. This study aimed at investigating whether PTBP1 expression associates with clinical outcome in patients with NMIBC. Experimental Design: A cohort of 152 patients presenting with primary NMIBC (pTa-pT1) was enrolled. Primary NMIBCs were assessed for PTBP1 expression by IHC, and the results were correlated with clinical data using Kaplan-Meier curves and Cox regression analyses. Cell proliferation and survival assays were performed to assess the function of PTBP1. Furthermore, the impact of PTBP1 on the AS pattern of specific bladder cancer-related genes was investigated in cancer cell lines and in patients' specimens. Results: Public datasets querying highlighted a positive correlation between PTBP1 expression and NMIBC progression, which was then confirmed by IHC analysis. High PTBP1 expression was associated with worse clinical outcome in terms of incidence of tumor relapse and survival in patients with NMIBC. Interestingly, downregulation of PTBP1 in bladder cancer cell lines affected prosurvival features. Accordingly, PTBP1 modulated AS of bladder cancer-related genes in cell lines and patient's specimens. Conclusions: PTBP1 expression correlates with disease progression, poor prognosis, and worse survival in patients with NMIBC. Downregulation of PTBP1 expression affects prosurvival features of bladder cancer cells and modulates AS of genes with relevance for bladder cancer, suggesting its role as an outcome-predictor in this disease. Clin Cancer Res; 24(21); 5422-32. ©2018 AACR ., (©2018 American Association for Cancer Research.)

Oropouche virus (OROV) is an emerging pathogen which causes Oropouche fever and meningitis in humans. Several outbreaks of OROV in South America, especially in Brazil, have changed its status as an emerging disease, but no vaccine or specific drug target is available yet. Our approach was to identify the epitope-based vaccine candidates as well as the ligand-binding pockets through the use of immunoinformatics. In this report, we identified both T-cell and B-cell epitopes of the most antigenic OROV polyprotein with the potential to induce both humoral and cell-mediated immunity. Eighteen highly antigenic and immunogenic CD8 + T-cell epitopes were identified, including three 100% conserved epitopes ( TSSWGCEEY , CSMCGLIHY , and LAIDTGCLY ) as the potential vaccine candidates. The selected epitopes showed 95.77% coverage for the mixed Brazilian population. The docking simulation ensured the binding interaction with high affinity. A total of five highly conserved and nontoxic linear B-cell epitopes " NQKIDLSQL ," " HPLSTSQIGDRC ," " SHCNLEFTAITADKIMSL ," " PEKIPAKEGWLTFSKEHTSSW ," and " HHYKPTKNLPHVVPRYH " were selected as potential vaccine candidates. The predicted eight conformational B-cell epitopes represent the accessibility for the entered virus. In the posttherapeutic strategy, ten ligand-binding pockets were identified for effective inhibitor design against emerging OROV infection. Collectively, this research provides novel candidates for epitope-based peptide vaccine design against OROV.

In the present study, an integrated approach combining HPLC/DAD, GC/MS, near infrared (NIR) spectroscopy, and chemometrics was used to geographically discriminate saffron samples from Iran and China. Chinese and Iranian samples can be well-separated based on HPLC data analysed by a principal component analysis and an orthogonal partial least squares discriminant analysis. Picrocrocin and two types of crocins were found to be the discriminating variables, and the Chinese samples had higher contents of safranal and picrocrocin but lower cis-crocin 3Gg, kaempferol-3-O-sophoroside and isophorone. Furthermore, an NIR method was successfully established to rapidly distinguish the Chinese and Iranian samples. The relationship between an ISO standard and the contents of the chemical indices was also studied. The results indicated that the ISO standard should be revised, especially for analysing safranal., (Copyright © 2018 Elsevier Ltd. All rights reserved.)

We have adapted the yeast 2-hybrid assay to simultaneously uncover dozens of transient and static protein interactions within a single screen utilizing high-throughput short-read DNA sequencing. The resulting sequence datasets can not only track what genes in a population that are enriched during selection for positive yeast 2-hybrid interactions, but also give detailed information about the relevant subdomains of proteins sufficient for interaction. Here, we describe a full suite of stand-alone software programs that allow non-experts to perform all the bioinformatics and statistical steps to process and analyze DNA sequence fastq files from a batch yeast 2-hybrid assay. The processing steps covered by these software include: 1) mapping and counting sequence reads corresponding to each candidate protein encoded within a yeast 2-hybrid prey library; 2) a statistical analysis program that evaluates the enrichment profiles; and 3) tools to examine the translational frame and position within the coding region of each enriched plasmid that encodes the interacting proteins of interest.

Nowadays, due to the advance of experimental techniques in glycomics, large collections of glycan profiles are regularly published. The rapid growth of available glycan data accentuates the lack of innovative tools for visualizing and exploring large amount of information. Scientists resort to using general-purpose spreadsheet applications to create ad hoc data visualization. Thus, results end up being encoded in publication images and text, while valuable curated data is stored in files as supplementary information. To tackle this problem, we have built an interactive pipeline composed with three tools: Glynsight, EpitopeXtractor and Glydin'. Glycan profile data can be imported in Glynsight, which generates a custom interactive glycan profile. Several profiles can be compared and glycan composition is integrated with structural data stored in databases. Glycan structures of interest can then be sent to EpitopeXtractor to perform a glycoepitope extraction. EpitopeXtractor results can be superimposed on the Glydin' glycoepitope network. The network visualization allows fast detection of clusters of glycoepitopes and discovery of potential new targets. Each of these tools is standalone or can be used in conjunction with the others, depending on the data and the specific interest of the user. All the tools composing this pipeline are part of the Glycomics@ExPASy initiative and are available at https://www.expasy.org/glycomics.

Prediction of compound properties from structure via quantitative structure-activity relationship and machine-learning approaches is an important computational chemistry task in small-molecule drug research. Though many such properties are dependent on three-dimensional structures or even conformer ensembles, the majority of models are based on descriptors derived from two-dimensional structures. Here we present results from a thorough benchmark study of force field, semiempirical, and density functional methods for the calculation of conformer energies in the gas phase and water solvation as a foundation for the correct identification of relevant low-energy conformers. We find that the tight-binding ansatz GFN-xTB shows the lowest error metrics and highest correlation to the benchmark PBE0-D3(BJ)/def2-TZVP in the gas phase for the computationally fast methods and that in solvent OPLS3 becomes comparable in performance. MMFF94, AM1, and DFTB+ perform worse, whereas the performance-optimized but far more expensive functional PBEh-3c yields energies almost perfectly correlated to the benchmark and should be used whenever affordable. On the basis of our findings, we have implemented a reliable and fast protocol for the identification of low-energy conformers of drug-like molecules in water that can be used for the quantification of strain energy and entropy contributions to target binding as well as for the derivation of conformer-ensemble-dependent molecular descriptors.

Making predictions with an associated confidence is highly desirable as it facilitates decision making and resource prioritization. Conformal regression is a machine learning framework that allows the user to define the required confidence and delivers predictions that are guaranteed to be correct to the selected extent. In this study, we apply conformal regression to model molecular properties and bioactivity values and investigate different ways to scale the resultant prediction intervals to create as efficient (i.e., narrow) regressors as possible. Different algorithms to estimate the prediction uncertainty were used to normalize the prediction ranges, and the different approaches were evaluated on 29 publicly available data sets. Our results show that the most efficient conformal regressors are obtained when using the natural exponential of the ensemble standard deviation from the underlying random forest to scale the prediction intervals, but other approaches were almost as efficient. This approach afforded an average prediction range of 1.65 pIC50 units at the 80% confidence level when applied to bioactivity modeling. The choice of nonconformity function has a pronounced impact on the average prediction range with a difference of close to one log unit in bioactivity between the tightest and widest prediction range. Overall, conformal regression is a robust approach to generate bioactivity predictions with associated confidence.

Most of the common molecular descriptors have numerous different implementations. This can influence the results of compound prioritization based on the multiparameter assessment (MPA) approach that allows a medicinal chemist to simultaneously analyze and achieve the desired balance of the diverse and often conflicting molecular and pharmacological properties. In this study, we analyzed the feasibility of using different implementations of common descriptors (logP, logS, TPSA, logBB, hERG, nHBA) interchangeably in predesigned sets of requirements in the course of multiparameter compound optimization. The influence of methods of descriptor calculation, continuity or discreteness of their values, their applicability domains, as well as of the nature of desirability functions in an MPA profile were examined in terms of the stability of MPA compound ranking. It was shown that the interchangeable use of different methods of descriptor calculation is reliably acceptable only for continuously distributed parameters transformed by a smooth desirability function. If a descriptor in an MPA scheme is discretely distributed, only the implementation that was used for building the scoring profile may be used for assessment. An inconsistency of assessment due to different applicability domains of descriptors was also demonstrated.