Utilizing Artificial Intelligence to Manage COVID-19 Scientific Evidence Torrent with Risklick AI: A Critical Tool for Pharmacology and Therapy Development.

INTRODUCTION: The SARS-CoV-2 pandemic has led to one of the most critical and boundless waves of publications in the history of modern science. The necessity to find and pursue relevant information and quantify its quality is broadly acknowledged. Modern information retrieval techniques combined with artificial intelligence (AI) appear as one of the key strategies for COVID-19 living evidence management. Nevertheless, most AI projects that retrieve COVID-19 literature still require manual tasks. METHODS: In this context, we pre-sent a novel, automated search platform, called Risklick AI, which aims to automatically gather COVID-19 scientific evidence and enables scientists, policy makers, and healthcare professionals to find the most relevant information tailored to their question of interest in real time. RESULTS: Here, we compare the capacity of Risklick AI to find COVID-19-related clinical trials and scientific publications in comparison with clinicaltrials.gov and PubMed in the field of pharmacology and clinical intervention. DISCUSSION: The results demonstrate that Risklick AI is able to find COVID-19 references more effectively, both in terms of precision and recall, compared to the baseline platforms. Hence, Risklick AI could become a useful alternative assistant to scientists fighting the COVID-19 pandemic.

Approaches to handling missing or "problematic" pharmacology data: Pharmacokinetics.

Missing or erroneous information is a common problem in the analysis of pharmacokinetic (PK) data. This may present as missing or inaccurate dose level or dose time, drug concentrations below the analytical limit of quantification, missing sample times, or missing or incorrect covariate information. Several methods to handle problematic data have been evaluated, although no single, broad set of recommendations for commonly occurring errors has been published. In this tutorial, we review the existing literature and present the results of our simulation studies that evaluated common methods to handle known data errors to bridge the remaining gaps and expand on the existing knowledge. This tutorial is intended for any scientist analyzing a PK data set with missing or apparently erroneous data. The approaches described herein may also be useful for the analysis of nonclinical PK data.

Using the IDEOM Workflow for LCMS-Based Metabolomics Studies of Drug Mechanisms.

Rapid advancements in metabolomics technologies have allowed for application of liquid chromatography mass spectrometry (LCMS)-based metabolomics to investigate a wide range of biological questions. In addition to an important role in studies of cellular biochemistry and biomarker discovery, an exciting application of metabolomics is the elucidation of mechanisms of drug action (Creek et al., Antimicrob Agents Chemother 60:6650-6663, 2016; Allman et al., Antimicrob Agents Chemother 60:6635-6649, 2016). Although it is a very useful technique, challenges in raw data processing, extracting useful information out of large noisy datasets, and identifying metabolites with confidence, have meant that metabolomics is still perceived as a highly specialized technology. As a result, metabolomics has not yet achieved the anticipated extent of uptake in laboratories around the world as genomics or transcriptomics. With a view to bring metabolomics within reach of a nonspecialist scientist, here we describe a routine workflow with IDEOM, which is a graphical user interface within Microsoft Excel, which almost all researchers are familiar with. IDEOM consists of custom built algorithms that allow LCMS data processing, automatic noise filtering and identification of metabolite features (Creek et al., Bioinformatics 28:1048-1049, 2012). Its automated interface incorporates advanced LCMS data processing tools, mzMatch and XCMS, and requires R for complete functionality. IDEOM is freely available for all researchers and this chapter will focus on describing the IDEOM workflow for the nonspecialist researcher in the context of studies designed to elucidate mechanisms of drug action.

[Roles and impacts of pharmaceutical activity from 1990 to today: Literature review and research perspectives]. / Rôles et impacts des interventions pharmaceutiques de 1990 à nos jours : revue de littérature et perspective de recherche.

CONTEXT: There is more and more evidence about the roles and impacts of the pharmacist. Health decision makers, managers, clinicians and patients need evidence to support an appropriate allocation of funds to different models of practice. OBJECTIVES: The main objective is to present an inventory of the roles and impacts of pharmaceutical activity in the international literature. METHODS: Review of literature. The articles related to the pharmacist's roles and impacts were selected according to a reproducible research strategy from 1990 to the present day (French/English with description of the intervention and impacts, where applicable) and a standard operating procedure. The following variables were extracted: author, country, specifications, pharmaceutical activities, care programs, targeted pathologies, impacts according to eight markers (mortality, morbidity, costs, adverse events, medication errors, compliance, satisfaction, others) and quality score. Only descriptive statistics were performed. RESULTS: As of February 1st, 2019, we recorded 2424 articles divided into 100 subjects (41 pharmaceutical activities, 30 pathologies, 29 care programs). Studies come from the United States (46.66%), multiple countries (8.00%), Canada (7.67%), France (6.06%), the United Kingdom (5.19%), Australia (3.50%) and other countries (22.92%). Studies are cross-sectional (47.55%), retrospective (33.68%) and prospective (17.87%) or non-categorized (<1%). The markers associated with the pharmacist's activity concern morbidity (23.12%), medication errors (11.82%), satisfaction (7.13%), compliance (6.06%), costs (5.47%), adverse events (3.74%), mortality (1.36%), and other indicators (41.31%). The studies have 6763 descriptive parameters and 5224 impact parameters (60.42% are positive, 38.55% are neutral and 1.03% are negative). CONCLUSION: This literature review confirms the roles and impacts of the pharmaceutical activity both in the pharmacy and in the hospital. A majority of the pharmaceutical interventions studied have positive impacts. It is essential to consider evidence about the roles and impacts of the pharmaceutical activities to take full advantage of the pharmacist's expertise in healthcare.

Evaluating the Pharmacological Management of Terminal Delirium in Imminently Dying Patients With and Without the Comfort Measure Order Set.

Terminal delirium is a distressing irreversible process that occurs frequently in the dying phase, often misdiagnosed and undertreated. A previous study in our organization revealed that terminal delirium was a poorly managed symptom at end of life. Pharmacological options are available in an existing order set to manage this symptom. The management plans of 41 patients identified as having terminal delirium were further evaluated. Elements extracted included medications prescribed to manage terminal delirium, whether medication changes occurred, and whether they were administered and effective. Patients with the order set were more comfortable as compared with the group without. Both groups had several changes made by the palliative care team. Nurses did not administer prescribed as-needed medication to more than one-third of patients. Modifications will be made to the existing order set, and additional education for staff will be organized.

An Invitation to Pharmacostatics.

Pharmacology, the study of interactions between biological processes and therapeutic agents, is traditionally presented as consisting of two subdisciplines: pharmacokinetics, which is about the distribution and metabolism of drugs in organisms, and pharmacodynamics, which is about the organisms' response to drugs. In discovery-stage pharmacology however, one primary concern is what we call pharmacostatics, the characterization of equilibrium parameters and states of core interactions of physiologic and therapeutic interest. This usually takes the form of studying dose-response curves, without consideration for the relevant qualitative properties of the underlying reaction networks, e.g., the existence, multiplicity and asymptotic stability of steady states. Furthermore, steady-state calculations customarily employ manually derived closed-form expressions based on approximating assumptions. While these formulas may seem adequate most of the time, the assumptions need not apply, and there are genuine though seemingly uncommon cases where this approach is not feasible and/or fails to explain non-monotone dose-response curves. It is this paper's aim to stimulate interest in mathematical problems arising in pharmacostatics. We specifically pose two problems about a particular relevant class of networks of reversible binding reactions. The first problem is to exploit a certain fixed-point formulation of the equilibrium equation to devise an algorithmic method that would be compellingly preferable to current practice in the pharmacostatics context. The second problem is to explicitly anticipate the possibility of non-monotone dose-response curves from network topology. Addressing these problems would positively impact biopharmaceutical research, and they have inherent mathematical interest.

Evaluation of FOCEI and SAEM Estimation Methods in Population Pharmacokinetic Analysis Using NONMEM® Across Rich, Medium, and Sparse Sampling Data.

BACKGROUND AND OBJECTIVES: First-order conditional estimation with interaction (FOCEI) is one of the most commonly used estimation methods in nonlinear mixed effects modeling, while the stochastic approximation expectation maximization (SAEM) is the newer estimation algorithm. This work aimed to compare the performance of FOCEI and SAEM methods when using NONMEM® with the classical one- and two-compartment models across rich, medium, and sparse data. METHODS: One- and two-compartment models of the previous studies were used to simulate data in three scenarios: rich, medium, and sparse data. For each scenario, there were 100 data sets, containing 100 individuals in each data set. Every data set was estimated with both FOCEI and SAEM methods. The simulation and estimation were performed using NONMEM®. The completion rates, percentage of relative estimation errors (%RERs), root mean square errors (RMSEs), and runtimes were considered to assess the completion, accuracy, precision, and speed of estimation, respectively. RESULTS: Both FOCEI and SAEM methods provided comparable completion rates, median %RERs (ranged from - 9.03 to 3.27% for FOCEI and - 9.17 to 3.27% for SAEM) and RMSEs (ranged from 0.0004 to 1.244 for FOCEI and 0.0004 to 1.131 for SAEM) for most parameters in both models across three scenarios. The run times were much shorter with FOCEI (ranged from 0.18 to 0.98 min) compared to SAEM method (ranged from 4.64 to 12.03 min). CONCLUSIONS: For the classical one- and two-compartment models, FOCEI method exhibited comparable performance similar to SAEM method but with significantly shorter runtimes across rich, medium, and sparse sampling scenarios.

Learning Medicinal Chemistry Absorption, Distribution, Metabolism, Excretion, and Toxicity (ADMET) Rules from Cross-Company Matched Molecular Pairs Analysis (MMPA).

The first large scale analysis of in vitro absorption, distribution, metabolism, excretion, and toxicity (ADMET) data shared across multiple major pharma has been performed. Using advanced matched molecular pair analysis (MMPA), we combined data from three pharmaceutical companies and generated ADMET rules, avoiding the need to disclose the full chemical structures. On top of the very large exchange of knowledge, all companies involved synergistically gained approximately 20% more rules from the shared transformations. There is good quantitative agreement between the rules based on shared data compared to both individual companies' rules and rules published in the literature. Known correlations between log  D, solubility, in vitro clearance, and plasma protein binding also hold in transformation space, but there are also interesting exceptions. Data pools such as this allow focusing on particular functional groups and characterizing their ADMET profile. Finally the role of a corpus of robustly tested medicinal chemistry knowledge in the training of medicinal chemistry is discussed.

Scientific Performance of Brazilian Researchers in Pharmacology with grants from CNPq: A comparative study within the Brazilian categories.

In Brazil, scientific performance of researchers is one important criteria for decision-making in grant allocation. In this context, this study aimed to evaluate and compare the profile of 82 seniors' investigators (graded as level 1A-D) which were receiving CNPq (National Council for Scientific and Technological Development) productivity grant in Pharmacology, by analyzing the pattern of citation of their papers and h-index. Total documents, citations (with and without self-citations) and h-index (with and without self-citations) were retrieved from the Scopus database. The results indicated a clear difference among researchers from the higher categories (1A and 1B) in most of the parameters analyzed. However, no noticeable differentiation was found between researchers from grant category 1C and 1D. The results presented here may inform the scientific community and the grant agencies on the profile of PQ 1(A-D) fellows of Pharmacology, and may help to define new differences within CNPq grant categories, and consequently, a better allocation of grants.

Approaches for modeling within subject variability in pharmacometric count data analysis: dynamic inter-occasion variability and stochastic differential equations.

Parameter variation in pharmacometric analysis studies can be characterized as within subject parameter variability (WSV) in pharmacometric models. WSV has previously been successfully modeled using inter-occasion variability (IOV), but also stochastic differential equations (SDEs). In this study, two approaches, dynamic inter-occasion variability (dIOV) and adapted stochastic differential equations, were proposed to investigate WSV in pharmacometric count data analysis. These approaches were applied to published count models for seizure counts and Likert pain scores. Both approaches improved the model fits significantly. In addition, stochastic simulation and estimation were used to explore further the capability of the two approaches to diagnose and improve models where existing WSV is not recognized. The results of simulations confirmed the gain in introducing WSV as dIOV and SDEs when parameters vary randomly over time. Further, the approaches were also informative as diagnostics of model misspecification, when parameters changed systematically over time but this was not recognized in the structural model. The proposed approaches in this study offer strategies to characterize WSV and are not restricted to count data.

Statistical power analysis of cardiovascular safety pharmacology studies in conscious rats.

UNLABELLED: Cardiovascular (CV) toxicity and related attrition are a major challenge for novel therapeutic entities and identifying CV liability early is critical for effective derisking. CV safety pharmacology studies in rats are a valuable tool for early investigation of CV risk. Thorough understanding of data analysis techniques and statistical power of these studies is currently lacking and is imperative for enabling sound decision-making. METHODS: Data from 24 crossover and 12 parallel design CV telemetry rat studies were used for statistical power calculations. Average values of telemetry parameters (heart rate, blood pressure, body temperature, and activity) were logged every 60s (from 1h predose to 24h post-dose) and reduced to 15min mean values. These data were subsequently binned into super intervals for statistical analysis. A repeated measure analysis of variance was used for statistical analysis of crossover studies and a repeated measure analysis of covariance was used for parallel studies. Statistical power analysis was performed to generate power curves and establish relationships between detectable CV (blood pressure and heart rate) changes and statistical power. Additionally, data from a crossover CV study with phentolamine at 4, 20 and 100mg/kg are reported as a representative example of data analysis methods. RESULTS: Phentolamine produced a CV profile characteristic of alpha adrenergic receptor antagonism, evidenced by a dose-dependent decrease in blood pressure and reflex tachycardia. Detectable blood pressure changes at 80% statistical power for crossover studies (n=8) were 4-5mmHg. For parallel studies (n=8), detectable changes at 80% power were 6-7mmHg. Detectable heart rate changes for both study designs were 20-22bpm. DISCUSSION: Based on our results, the conscious rat CV model is a sensitive tool to detect and mitigate CV risk in early safety studies. Furthermore, these results will enable informed selection of appropriate models and study design for early stage CV studies.

Evaluation of estimation, prediction and inference for autocorrelated latent variable modeling of binary data-a simulation study.

Longitudinal models of binary or ordered categorical data are often evaluated for adequacy by the ability of these to characterize the transition frequency and type between response states. Drug development decisions are often concerned with accurate prediction and inference of the probability of response by time and dose. A question arises on whether the transition probabilities need to be characterized adequately to ensure accurate response prediction probabilities unconditional on the previous response state. To address this, a simulation study was conducted to assess bias in estimation, prediction and inferences of autocorrelated latent variable models (ALVMs) when the transition probabilities are misspecified due to ill-posed random effects structures, inadequate likelihood approximation or omission of the autocorrelation component. The results may be surprising in that these suggest that characterizing autocorrelation in ALVMs is not as important as specifying a suitably rich random effects structure.

The analysis of dose-response curve from bioassays with quantal response: Deterministic or statistical approaches?

Dose-response relations can be obtained from systems at any structural level of biological matter, from the molecular to the organismic level. There are two types of approaches for analyzing dose-response curves: a deterministic approach, based on the law of mass action, and a statistical approach, based on the assumed probabilities distribution of phenotypic characters. Models based on the law of mass action have been proposed to analyze dose-response relations across the entire range of biological systems. The purpose of this paper is to discuss the principles that determine the dose-response relations. Dose-response curves of simple systems are the result of chemical interactions between reacting molecules, and therefore are supported by the law of mass action. In consequence, the shape of these curves is perfectly sustained by physicochemical features. However, dose-response curves of bioassays with quantal response are not explained by the simple collision of molecules but by phenotypic variations among individuals and can be interpreted as individual tolerances. The expression of tolerance is the result of many genetic and environmental factors and thus can be considered a random variable. In consequence, the shape of its associated dose-response curve has no physicochemical bearings; instead, they are originated from random biological variations. Due to the randomness of tolerance there is no reason to use deterministic equations for its analysis; on the contrary, statistical models are the appropriate tools for analyzing these dose-response relations.

The ISoP Standards and Best Practices Committee.

The mission of the International Society of Pharmacometrics (ISoP) Standards and Best Practices Committee is to provide best practices and recommendations for standard pharmacometric analyses-e.g., population pharmacokinetics/pharmacodynamics (PK/PD), exposure-response, disease models-with the goal of increasing consistency, productivity, quality, communication, and impact of pharmacometrics on decision making. We present the progress and plans of the committee and call for volunteers to start new initiatives.

The use and misuse of statistical methodologies in pharmacology research.

Descriptive, exploratory, and inferential statistics are necessary components of hypothesis-driven biomedical research. Despite the ubiquitous need for these tools, the emphasis on statistical methods in pharmacology has become dominated by inferential methods often chosen more by the availability of user-friendly software than by any understanding of the data set or the critical assumptions of the statistical tests. Such frank misuse of statistical methodology and the quest to reach the mystical α<0.05 criteria has hampered research via the publication of incorrect analysis driven by rudimentary statistical training. Perhaps more critically, a poor understanding of statistical tools limits the conclusions that may be drawn from a study by divorcing the investigator from their own data. The net result is a decrease in quality and confidence in research findings, fueling recent controversies over the reproducibility of high profile findings and effects that appear to diminish over time. The recent development of "omics" approaches leading to the production of massive higher dimensional data sets has amplified these issues making it clear that new approaches are needed to appropriately and effectively mine this type of data. Unfortunately, statistical education in the field has not kept pace. This commentary provides a foundation for an intuitive understanding of statistics that fosters an exploratory approach and an appreciation for the assumptions of various statistical tests that hopefully will increase the correct use of statistics, the application of exploratory data analysis, and the use of statistical study design, with the goal of increasing reproducibility and confidence in the literature.