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

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

Blinded Predictions and Post Hoc Analysis of the Second Solubility Challenge Data: Exploring Training Data and Feature Set Selection for Machine and Deep Learning Models.

Accurate methods to predict solubility from molecular structure are highly sought after in the chemical sciences. To assess the state of the art, the American Chemical Society organized a "Second Solubility Challenge" in 2019, in which competitors were invited to submit blinded predictions of the solubilities of 132 drug-like molecules. In the first part of this article, we describe the development of two models that were submitted to the Blind Challenge in 2019 but which have not previously been reported. These models were based on computationally inexpensive molecular descriptors and traditional machine learning algorithms and were trained on a relatively small data set of 300 molecules. In the second part of the article, to test the hypothesis that predictions would improve with more advanced algorithms and higher volumes of training data, we compare these original predictions with those made after the deadline using deep learning models trained on larger solubility data sets consisting of 2999 and 5697 molecules. The results show that there are several algorithms that are able to obtain near state-of-the-art performance on the solubility challenge data sets, with the best model, a graph convolutional neural network, resulting in an RMSE of 0.86 log units. Critical analysis of the models reveals systematic differences between the performance of models using certain feature sets and training data sets. The results suggest that careful selection of high quality training data from relevant regions of chemical space is critical for prediction accuracy but that other methodological issues remain problematic for machine learning solubility models, such as the difficulty in modeling complex chemical spaces from sparse training data sets.

Disparities in health care use among low-salary and high-salary employees.

OBJECTIVES: To examine how health care utilization and spending vary for low-income employees compared with high-income employees enrolled in an employer-sponsored high-deductible health plan (HDHP). STUDY DESIGN: We use commercial medical claims data and administrative human resource data from a large employer between 2014 and 2018. We link the administrative data, which include details on salary and other benefit choices, to each employee in each year with medical claims. Our variables of interest include medical spending and utilization outcomes grouped into different care settings. METHODS: Using multivariate regressions, we estimate the association between salary buckets and health care utilization and spending, controlling for demographic characteristics, comorbidities of employees, human resource health plan benefits, and geography. RESULTS: Employees earning less than $75,000 show lower rates of utilization and spending on preventive measures, such as outpatient visits and prescription drugs, while having higher rates of utilization of preventable and avoidable emergency department visits and inpatient stays, resulting in lower overall health care spending among lower-salary employees. CONCLUSIONS: Low-salary employees enrolled in HDHPs have higher rates of acute care utilization and spending but lower rates of primary care spending compared with high-salary employees. Results suggest that HDHPs discourage routine physician-patient care among low-salary employees.

Alchemical Free Energy Methods Applied to Complexes of the First Bromodomain of BRD4.

Accurate and rapid predictions of the binding affinity of a compound to a target are one of the ultimate goals of computer aided drug design. Alchemical approaches to free energy estimations follow the path from an initial state of the system to the final state through alchemical changes of the energy function during a molecular dynamics simulation. Herein, we explore the accuracy and efficiency of two such techniques: relative free energy perturbation (FEP) and multisite lambda dynamics (MSλD). These are applied to a series of inhibitors for the bromodomain-containing protein 4 (BRD4). We demonstrate a procedure for obtaining accurate relative binding free energies using MSλD when dealing with a change in the net charge of the ligand. This resulted in an impressive comparison with experiment, with an average difference of 0.4 ± 0.4 kcal mol-1. In a benchmarking study for the relative FEP calculations, we found that using 20 lambda windows with 0.5 ns of equilibration and 1 ns of data collection for each window gave the optimal compromise between accuracy and speed. Overall, relative FEP and MSλD predicted binding free energies with comparable accuracy, an average of 0.6 kcal mol-1 for each method. However, MSλD makes predictions for a larger molecular space over a much shorter time scale than relative FEP, with MSλD requiring a factor of 18 times less simulation time for the entire molecule space.

Is There an Optimal Antiplatelet Strategy after Gastrointestinal Bleeding in Patients with Coronary Artery Disease?

Gastrointestinal bleeding after percutaneous coronary intervention (PCI) is a not too uncommon clinical situation and is associated with high morbidity and mortality. After initial treatment, a number of clinical decisions must be made weighing the risks of ischemic events and future bleeding. In particular, healthcare providers must carefully balance the effectiveness of antiplatelet therapy in the secondary prevention of coronary events, primarily future spontaneous myocardial infarction and stent thrombosis, against the risk of major, most commonly gastrointestinal bleeding. The first question is whether a dual antiplatelet therapy strategy is required or if a single antiplatelet agent will suffice. Then, if a single antiplatelet agent is adequate, which agent should be continued. Although there is some guidance to answer some of these questions, there are inadequate evidence-based data for others. Below, we review the various considerations and summarize our approach and rationale to manage patients who had gastrointestinal bleeding after PCI.

Structural variation of protein-ligand complexes of the first bromodomain of BRD4.

The bromodomain-containing protein 4 (BRD4), a member of the bromodomain and extra-terminal domain (BET) family, plays a key role in several diseases, especially cancers. With increased interest in BRD4 as a therapeutic target, many X-ray crystal structures of the protein in complex with small molecule inhibitors are publicly available over the recent decade. In this study, we use this structural information to investigate the conformations of the first bromodomain (BD1) of BRD4. Structural alignment of 297 BRD4-BD1 complexes shows a high level of similarity between the structures of BRD4-BD1, regardless of the bound ligand. We employ WONKA, a tool for detailed analyses of protein binding sites, to compare the active site of over 100 of these crystal structures. The positions of key binding site residues show a high level of conformational similarity, with the exception of Trp81. A focused analysis on the highly conserved water network in the binding site of BRD4-BD1 is performed to identify the positions of these water molecules across the crystal structures. The importance of the water network is illustrated using molecular docking and absolute free energy perturbation simulations. 82% of the ligand poses were better predicted when including water molecules as part of the receptor. Our analysis provides guidance for the design of new BRD4-BD1 inhibitors and the selection of the best structure of BRD4-BD1 to use in structure-based drug design, an important approach for faster and more cost-efficient lead discovery.

A Turing Test for Molecular Generators.

Machine learning approaches promise to accelerate and improve success rates in medicinal chemistry programs by more effectively leveraging available data to guide a molecular design. A key step of an automated computational design algorithm is molecule generation, where the machine is required to design high-quality, drug-like molecules within the appropriate chemical space. Many algorithms have been proposed for molecular generation; however, a challenge is how to assess the validity of the resulting molecules. Here, we report three Turing-inspired tests designed to evaluate the performance of molecular generators. Profound differences were observed between the performance of molecule generators in these tests, highlighting the importance of selection of the appropriate design algorithms for specific circumstances. One molecule generator, based on match molecular pairs, performed excellently against all tests and thus provides a valuable component for machine-driven medicinal chemistry design workflows.

Shift Work Sleep Disorder and Night Shift Work Significantly Impair Erectile Function.

OBJECTIVES: Here we examine the association between shift work sleep disorder (SWSD) and erectile dysfunction (ED) in shift workers. METHODS: Men presenting to a single andrology clinic between January 2014 and July 2017 completed validated questionnaires: International Index of Erectile Function (IIEF), Patient Health Questionnaire-9 (PHQ-9), and the nonvalidated SWSD Questionnaire. Men were also asked about shift work schedule, comorbidities, phosphodiesterase 5 (PDE5) inhibitor use, and testosterone use. Serum total testosterone values were determined for each visit. Linear regression was performed controlling for testosterone use, testosterone levels, PDE5 inhibitor use, age, and comorbidities to determine the effect of SWSD on ED as assessed using the IIEF. RESULTS: Of the 754 men completing questionnaires, 204 reported nonstandard shift work (begins before 7 am or after 6 pm, regularly extends out of that frame, or rotates frequently) and 48 were found to have SWSD using a screening questionnaire. Nonstandard shift work alone did not result in worse IIEF-EF scores (P = .31), but those who worked nonstandard shifts and had SWSD demonstrated IIEF-EF scores 2.8 points lower than men without SWSD (P < .01). When assessing for the type of shift work performed, men who worked night shifts had IIEF-EF scores 7.6 points lower than men who worked during the day or evening (P < .01). Testosterone use improved IIEF-EF scores for men with SWSD by 2.9 points, ameliorating the effect of SWSD on ED. However, baseline testosterone levels were not associated with worse erectile function in this cohort. CONCLUSION: Men with SWSD have worse erectile function, with men who work night shifts having even poorer erectile function. These findings suggest that circadian rhythm disturbance may significantly impact erectile function. While testosterone therapy may partly reverse the effects of SWSD, shift work is a potential risk factor for ED and should be assessed for as part of the evaluation of men with ED. Rodriguez KM, Kohn TP, Kohn JR, et al. Shift Work Sleep Disorder and Night Shift Work Significantly Impair Erectile Function. J Sex Med 2020;17:1687-1693.

Guidelines for Recurrent Neural Network Transfer Learning-Based Molecular Generation of Focused Libraries.

Deep learning approaches have become popular in recent years in the field of de novo molecular design. While a variety of different methods are available, it is still a challenge to assess and compare their performance. A particularly promising approach for automated drug design is to use recurrent neural networks (RNNs) as SMILES generators and train them with the learning procedure called "transfer learning". This involves first training the initial model on a large generic data set of molecules to learn the general syntax of SMILES, followed by fine-tuning on a smaller set of molecules, coming from, e.g., a lead optimization program. To create a well-performing transfer learning application which can be automated, it is important to understand how the size of the second data set affects the training process. In addition, extensive postfiltering using similarity metrics of the molecules generated after transfer learning should be avoided, as it can introduce new biases toward the selection of drug candidates. Here, we present results from the application of a gated recurrent unit cell (GRU)-RNN to transfer learning on data sets of varying sizes and complexity. Analysis of the results has allowed us to provide some general guidelines for transfer learning. In particular, we show that data set sizes containing at least 190 molecules are needed for effective GRU-RNN-based molecular generation using transfer learning. The methods presented here should be applicable generally to the benchmarking of other deep learning methodologies for molecule generation.

Increased Risk of Hypogonadal Symptoms in Shift Workers With Shift Work Sleep Disorder.

OBJECTIVE: To examine the association between shift work sleep disorder (SWSD), a primary circadian rhythm disorder characterized by excessive day-time sleepiness associated with shift work, and hypogonadal symptoms in shift workers. METHODS: Men presenting to an andrology clinic between July 2014 and June 2017 completed questionnaires assessing shift work schedule, SWSD risk, and hypogonadal symptoms ([quantitative] Androgen Deficiency in the Aging Male [qADAM, ADAM]). The impact of nonstandard shift work and SWSD on responses to qADAM and ADAM was assessed using ANOVA and linear regression. RESULTS: About 24.1% (619/2571) of men worked nonstandard shifts. Of those, 196 (31.7%) were considered to have SWSD. Controlling for age, comorbidities, and testosterone (T) levels, nonstandard shift workers had qADAM scores 1.12 points lower than day-time workers (P <.01). Subgroup analysis of nonstandard shift workers showed that those with SWSD had qADAM scores 5.47 points lower than men without SWSD (P <.01). In this same subgroup analysis, SWSD was independently associated with lower T levels (mean decrease 100.4 ng/dL, P <.01) when controlling for age, comorbidities, and prior T supplementation. CONCLUSION: Nonstandard shift workers with SWSD have even worse hypogonadal symptoms and lower T levels than day-time workers and nonstandard shift workers without SWSD. This suggests that poor sleep habits, as identified by SWSD, may contribute to the more severe hypogonadal symptoms seen in nonstandard shift workers.

Correction to: BRADSHAW: a system for automated molecular design.

The original version of this article unfortunately contained some mistakes in the references.

BRADSHAW: a system for automated molecular design.

This paper introduces BRADSHAW (Biological Response Analysis and Design System using an Heterogenous, Automated Workflow), a system for automated molecular design which integrates methods for chemical structure generation, experimental design, active learning and cheminformatics tools. The simple user interface is designed to facilitate access to large scale automated design whilst minimising software development required to introduce new algorithms, a critical requirement in what is a very fast moving field. The system embodies a philosophy of automation, best practice, experimental design and the use of both traditional cheminformatics and modern machine learning algorithms.

Shift Workers With Shift Work Disorder Have Worse Lower Urinary Tract Symptoms.

OBJECTIVE: To examine the association between shift work or shift work disorder (SWD) and lower urinary tract symptoms (LUTS). Nonstandard shift workers are defined as those working shifts outside of a normal 7 AM-6 PM work day. METHODS: Men presenting to a single andrology clinic between July 2014 and June 2017 completed questionnaires that included questions about work schedules, shift work status, SWD[1][1], personal well-being via the Patient Health Questionnaire-9, and LUTS (International Prostate Symptom Score [IPSS]). Men who had previously undergone prostate surgery were excluded. Shift work and SWD on IPSS was assessed via multivariate linear regression. RESULTS: Of the 2571 men who completed all questionnaires, 619 (24.1%) reported working nonstandard shifts in the past month. Of these, 196 (31.7%) had high risk of SWD as determined by a questionnaire. When controlling for age, medications, surgical intervention for benign prostatic hyperplasia, comorbidities, and testosterone (T) levels, nonstandard shift work overall was not associated with worse LUTS (P = .82). However, nonstandard shift workers at high risk for SWD had IPSS scores that were clinically significantly higher (3.74 points ± 0.57 standard error) than nonstandard shift workers without SWD (P <.0001). CONCLUSION: Nonstandard shift workers at high risk for SWD have worse LUTS than those without SWD. However, no association between nonstandard shift work and LUTS was found.

De Novo Molecule Design by Translating from Reduced Graphs to SMILES.

A key component of automated molecular design is the generation of compound ideas for subsequent filtering and assessment. Recently deep learning approaches have been explored as alternatives to traditional de novo molecular design techniques. Deep learning algorithms rely on learning from large pools of molecules represented as molecular graphs (generally SMILES), and several approaches can be used to tailor the generated molecules to defined regions of chemical space. Cheminformatics has developed alternative higher-level representations that capture the key properties of a set of molecules, and it would be of interest to understand whether such representations can be used to constrain the output of molecule generation algorithms. In this work we explore the use of one such representation, the Reduced Graph, as a definition of target chemical space for a deep learning molecule generator. The Reduced Graph replaces functional groups with superatoms representing the pharmacophoric features. Assigning these superatoms to specific nonorganic element types allows the Reduced Graph to be represented as a valid SMILES string. The mapping from standard SMILES to Reduced Graph SMILES is well-defined, however, the inverse is not true, and this presents a particular challenge. Here we present the results of a novel seq-to-seq approach to molecule generation, where the one to many mapping of Reduced Graph to SMILES is learned on a large training set. This training needs to be performed only once. In a subsequent step, this model can be used to generate arbitrary numbers of compounds that have the same Reduced Graph as any input molecule. Through analysis of data sets in ChEMBL we show that the approach generates valid molecules and can extrapolate to Reduced Graphs unseen in the training set. The method offers an alternative deep learning approach to molecule generation that does not rely on transfer learning, latent space generation, or adversarial networks and is applicable to scaffold hopping and other cheminformatics applications in drug discovery.

Nuisance Compounds, PAINS Filters, and Dark Chemical Matter in the GSK HTS Collection.

High-throughput screening (HTS) hits include compounds with undesirable properties. Many filters have been described to identify such hits. Notably, pan-assay interference compounds (PAINS) has been adopted by the community as the standard term to refer to such filters, and very useful guidelines have been adopted by the American Chemical Society (ACS) and subsequently triggered a healthy scientific debate about the pitfalls of draconian use of filters. Using an inhibitory frequency index, we have analyzed in detail the promiscuity profile of the whole GlaxoSmithKline (GSK) HTS collection comprising more than 2 million unique compounds that have been tested in hundreds of screening assays. We provide a comprehensive analysis of many previously published filters and newly described classes of nuisance structures that may serve as a useful source of empirical information to guide the design or growth of HTS collections and hit triaging strategies.

Advances in the management of dyslipidemia.

PURPOSE OF REVIEW: Cardiovascular disease is the leading cause of morbidity and mortality in the United States and therapies aimed at lipid modification are important for the reduction of cardiovascular risk. There have been many exciting advances in lipid management over the recent years. This review discusses these recent advances as well as the direction of future studies. RECENT FINDINGS: Several recent clinical trials support low-density lipoprotein cholesterol (LDL-c) reduction beyond maximal statin therapy for improved cardiovascular outcomes. Ezetimibe reduced LDL-c beyond maximal statin therapy and was associated with improved cardiovascular outcomes for high-risk populations. Further LDL-c reduction may also be achieved with proprotein convertase subtilisin/kexin type-9 (PCSK9) inhibition and a recent trial, Further Cardiovascular Outcomes Research with PCSK9 Inhibition in Subjects with Elevated Risk (FOURIER), was the first to show reduction in cardiovascular events for evolocumab. Additional outcome studies of monoclonal antibody and RNA-targeted PCSK9 inhibitors are underway. Quantitative high-density lipoprotein cholesterol (HDL-c) improvements have failed to have clinical impact to date; most recently, cholesteryl ester transfer protein inhibitors and apolipoprotein infusions have demonstrated disappointing results. There are still ongoing trials in both of these areas, but some newer therapies are focusing on HDL functionality and not just the absolute HDL-c levels. There are several ongoing studies in triglyceride reduction including fatty acid therapy, inhibition of apolipoprotein C-3 or ANGTPL3 and peroxisome proliferator-activated receptor-α agonists. SUMMARY: Lipid management continues to evolve and these advances have the potential to change clinical practice in the coming years.

Gain in Insurance Coverage and Residual Uninsurance Under the Affordable Care Act: Texas, 2013-2016.

OBJECTIVES: To examine the effects of the Affordable Care Act's (ACA's) Marketplace on Texas residents and determine which population subgroups benefited the most and which the least. METHODS: We analyzed insurance coverage rates among nonelderly Texas adults using the Health Reform Monitoring Survey-Texas from September 2013, just before the first open enrollment period in the Marketplace, through March 2016. RESULTS: Texas has experienced a roughly 6-percentage-point increase in insurance coverage (from 74.7% to 80.6%; P = .012) after implementation of the major insurance provisions of the ACA. The 4 subgroups with the largest increases in adjusted insurance coverage between 2013 and 2016 were persons aged 50 to 64 years (12.1 percentage points; P = .002), Hispanics (10.9 percentage points; P = .002), persons reporting fair or poor health status (10.2 percentage points; P = .038), and those with a high school diploma as their highest educational attainment (9.2 percentage points; P = .023). CONCLUSIONS: Many population subgroups have benefited from the ACA's Marketplace, but approximately 3 million Texas residents still lack health coverage. Adopting the ACA's Medicaid expansion is a means to address the lack of coverage.

Age and duration of testosterone therapy predict time to return of sperm count after human chorionic gonadotropin therapy.

OBJECTIVE: To determine factors that influence sperm recovery after T-associated infertility. DESIGN: Clinical retrospective study. SETTING: Academic male-infertility urology clinic. PATIENT(S): Sixty-six men who presented with infertility after T use. INTERVENTION(S): T cessation and combination high-dose hCG and selective estrogen modulator (SERM) therapy. MAIN OUTCOME MEASURE(S): Whether patients successfully achieved or failed to achieve a total motile count (TMC) of greater than 5 million sperm within 12 months of T cessation and initiation of therapy. RESULT(S): A TMC of greater than 5 million sperm was achieved by 46 men (70%). Both increased age and duration of T use directly correlated with time to sperm recovery at both 6 and 12 months of hCG/SERM therapy. Age more consistently limited sperm recovery, while duration of T use had less influence at 12 months than at 6 months. Only 64.8% of azoospermic men achieved a TMC greater than 5 million sperm at 12 months, compared with 91.7% of cryptozoospermic men, yet this did not predict a failure of sperm recovery. CONCLUSION(S): Increasing age and duration of T use significantly reduce the likelihood of recovery of sperm in the ejaculate, based on a criterion of a TMC of 5 million sperm, at 6 and 12 months. Physicians should be cautious in pursuing long-term T therapy, particularly in men who still desire fertility. Using these findings, physicians can counsel men regarding the likelihood of recovery of sperm at 6 and 12 months.