Preclinical validation of therapeutic targets predicted by tensor factorization on heterogeneous graphs.

Incorrect drug target identification is a major obstacle in drug discovery. Only 15% of drugs advance from Phase II to approval, with ineffective targets accounting for over 50% of these failures1-3. Advances in data fusion and computational modeling have independently progressed towards addressing this issue. Here, we capitalize on both these approaches with Rosalind, a comprehensive gene prioritization method that combines heterogeneous knowledge graph construction with relational inference via tensor factorization to accurately predict disease-gene links. Rosalind demonstrates an increase in performance of 18%-50% over five comparable state-of-the-art algorithms. On historical data, Rosalind prospectively identifies 1 in 4 therapeutic relationships eventually proven true. Beyond efficacy, Rosalind is able to accurately predict clinical trial successes (75% recall at rank 200) and distinguish likely failures (74% recall at rank 200). Lastly, Rosalind predictions were experimentally tested in a patient-derived in-vitro assay for Rheumatoid arthritis (RA), which yielded 5 promising genes, one of which is unexplored in RA.

Impact of neuroanatomical variations and electrode orientation on stimulus current in a device for migraine: a computational study.

OBJECTIVE: Conventional treatment methods for migraine often have side effects. One treatment involves a wearable neuromodulator targeting frontal nerves. Studies based on this technique have shown limited efficacy and the existing setting can cause pain. These may be associated with neuroanatomical variations which lead to high levels of required stimulus current. The aim of this paper is to study the effect of such variations on the activation currents of the Cefaly neuromodulator. Also, using a different electrode orientation, the possibility of reducing activation current levels to avoid painful side-effects and improve efficacy, is explored. APPROACH: This paper investigates the effect of neuroanatomical variations and electrode orientation on the stimulus current thresholds using a computational hybrid model involving a volume conductor and an advanced nerve model. Ten human head models are developed considering statistical variations of key neuroanatomical features, to model a representative population. MAIN RESULTS: By simulating the required stimulus current level in the head models, it is shown that neuroanatomical variations have a significant impact on the outcome, which is not solely a function of one specific neuroanatomical feature. The stimulus current thresholds based on the conventional Cefaly system vary from 4.4 mA to 25.1 mA across all head models. By altering the electrode orientation to align with the nerve branches, the stimulus current thresholds are substantially reduced to between 0.28 mA and 15 mA, reducing current density near pain-sensitive structures which may lead to a higher level of patient acceptance, further improving the efficacy. SIGNIFICANCE: Computational modeling based on statistically valid neuroanatomical parameters, covering a representative adult population, offers a powerful tool for quantitative comparison of the effect of the position of stimulating electrodes which is otherwise not possible in clinical studies.

CiPA challenges and opportunities from a non-clinical, clinical and regulatory perspectives. An overview of the safety pharmacology scientific discussion.

The Safety Pharmacology Society organized a scientific session at its annual conference in 2017 to discuss the challenges and opportunities of the Comprehensive In-Vitro Proarrhythmia Assay (CiPA) paradigm. Our intention was to raise awareness of this initiative with its members and also to gauge the extent to which safety pharmacologists have incorporated the CiPA testing strategy within the pharmaceutical industry. CiPA offers many potential opportunities including 1) a focus on proarrhythmic risk (as opposed to QTc prolongation), 2) providing scientific rationale to support the continued development of compounds that may have a poor selectivity over hERG whilst also blocking other inward currents and 3) reducing the extent of ECG monitoring in clinical trials with a greater influence of the non-clinical studies. Such opportunities may speed drug development and reduce costs. However, there are also challenges for CiPA implementation. For example, the mixed ion channel paradigm does not easily lend itself to a prospective drug discovery strategy although testing for such effects can be achieved with assays with good throughput. However, it should also be recognized that compounds with a mixed ion channel profile might also have properties that are undesirable to treat non-life threatening indications. All components of CiPA (nonclinical and clinical) require validation, particularly as a composite package to impact drug development and evaluation. One of the significant discussion points was that the existing regulatory guidance supports the use of components of CiPA through follow-up studies. A survey of the conference audience showed that the level of awareness of CiPA is quite high and that companies are already conducting some testing against a wider panel of cardiac ion channels beyond hERG. However, the adoption of other technologies (stem cell derived cardiac myocytes and in silico modeling) is less well developed. Taken together, the session demonstrated the potential advantages of CiPA, but also some significant challenges.

Pharmacokinetics of meropenem in children receiving continuous renal replacement therapy: Validation of clinical trial simulations.

Meropenem is frequently prescribed in critically ill children receiving continuous renal replacement therapy (CRRT). We previously used clinical trial simulations to evaluate dosing regimens of meropenem in this population and reported that a dose of 20 mg/kg every 12 hours optimizes target attainment. Meropenem pharmacokinetics were investigated in this prospective, open-label study to validate our previous in silico predictions. Seven patients received meropenem (13.8-22 mg/kg) administered intravenously every 12 hours as part of standard care. A mean dose of 18.6 mg/kg of meropenem was administered, resulting in a mean peak concentration of 80.1 µg/mL. Meropenem volume of distribution was 0.35 ± 0.085 L/kg. CRRT clearance was 40.2 ± 6.6 mL/(min · 1.73 m(2) ) and accounted for 63.4% of the total clearance of 74.8 ± 36.9 mL/(min · 1.73 m(2) ). Simulations demonstrated that a dose of 20 mg/kg every 12 hours resulted in a time above the minimum inhibitory concentration (%fT > MIC) of 100% in 5 out of 7 subjects, with a %fT > MIC of 93% and 43% in the remaining 2 subjects. We conclude that CRRT contributed significantly to the total clearance of meropenem. A dosing regimen of 20 mg/kg achieved good target attainment in critically ill children receiving CRRT, which is consistent with our previously published in silico predictions.

Full immersion simulation: validation of a distributed simulation environment for technical and non-technical skills training in Urology.

OBJECTIVE: To evaluate the face, content and construct validity of the distributed simulation (DS) environment for technical and non-technical skills training in endourology. To evaluate the educational impact of DS for urology training. SUBJECTS AND METHODS: DS offers a portable, low-cost simulated operating room environment that can be set up in any open space. A prospective mixed methods design using established validation methodology was conducted in this simulated environment with 10 experienced and 10 trainee urologists. All participants performed a simulated prostate resection in the DS environment. Outcome measures included surveys to evaluate the DS, as well as comparative analyses of experienced and trainee urologist's performance using real-time and 'blinded' video analysis and validated performance metrics. Non-parametric statistical methods were used to compare differences between groups. RESULTS: The DS environment demonstrated face, content and construct validity for both non-technical and technical skills. Kirkpatrick level 1 evidence for the educational impact of the DS environment was shown. Further studies are needed to evaluate the effect of simulated operating room training on real operating room performance. CONCLUSIONS: This study has shown the validity of the DS environment for non-technical, as well as technical skills training. DS-based simulation appears to be a valuable addition to traditional classroom-based simulation training.

Computational models of upper-limb motion during functional reaching tasks for application in FES-based stroke rehabilitation.

Functional electrical stimulation (FES) has been shown to be an effective approach to upper-limb stroke rehabilitation, where it is used to assist arm and shoulder motion. Model-based FES controllers have recently confirmed significant potential to improve accuracy of functional reaching tasks, but they typically require a reference trajectory to track. Few upper-limb FES control schemes embed a computational model of the task; however, this is critical to ensure the controller reinforces the intended movement with high accuracy. This paper derives computational motor control models of functional tasks that can be directly embedded in real-time FES control schemes, removing the need for a predefined reference trajectory. Dynamic models of the electrically stimulated arm are first derived, and constrained optimisation problems are formulated to encapsulate common activities of daily living. These are solved using iterative algorithms, and results are compared with kinematic data from 12 subjects and found to fit closely (mean fitting between 63.2% and 84.0%). The optimisation is performed iteratively using kinematic variables and hence can be transformed into an iterative learning control algorithm by replacing simulation signals with experimental data. The approach is therefore capable of controlling FES in real time to assist tasks in a manner corresponding to unimpaired natural movement. By ensuring that assistance is aligned with voluntary intention, the controller hence maximises the potential effectiveness of future stroke rehabilitation trials.

Visual control strategies of surgeons: a novel method of establishing the construct validity of a transurethral resection of the prostate surgical simulator.

OBJECTIVE: To examine novice and expert differences in visual control strategies while performing a virtual reality transurethral resection of the prostate (TURP) task and to determine if these differences could provide a novel method for assessing construct validity of the simulator. SUBJECTS AND METHODS: A total of 11 novices (no TURP experience) and 7 experts (>200 TURPs) completed a virtual reality prostate resection task on the TURPsim (Simbionix USA Corp, Cleveland, OH) while wearing an eye tracker (ASL, Bedford, MA). Performance parameters and the surgeon's visual control strategy were measured and compared between the 2 groups. RESULTS: Experts resected a greater percentage of prostate than novices (p < 0.01) and had less active diathermy time without tissue contact (p < 0.01). Experts adopted a target-locking visual strategy, employing fewer visual fixations (p < 0.05) with longer mean fixation duration (p < 0.005). With multiple learning trials, novices' performance improved and the adoption of a more expertlike gaze strategy was observed. CONCLUSION: Significant differences between experts and novices in both performance and visual control strategy were observed. The study of visual control strategies may be a useful adjunct, alongside measurements of motor performance, providing a novel method of assessing the construct validity of surgical simulators.

Assessing visual control during simulated and live operations: gathering evidence for the content validity of simulation using eye movement metrics.

BACKGROUND: Although virtual reality (VR) simulators serve an important role in the training and assessment of surgeons, they need to be evaluated for evidence of validity. Eye-tracking technology and measures of visual control have been used as an adjunct to the performance parameters produced by VR simulators to help in objectively establishing the construct validity (experts vs. novices) of VR simulators. However, determining the extent to which VR simulators represent the real procedure and environment (content validity) has largely been a subjective process undertaken by experienced surgeons. This study aimed to examine the content validity of a VR transurethral resection of the prostate (TURP) simulator by comparing visual control metrics taken during simulated and real TURP procedures. METHODS: Eye-tracking data were collected from seven surgeons performing 14 simulated TURP operations and three surgeons performing 15 real TURP operations on live patients. The data were analyzed offline, and visual control metrics (number and duration of fixations, percentage of time the surgeons fixated on the screen) were calculated. RESULTS: The surgeons displayed more fixations of a shorter duration and spent less time fixating on the video monitor during the real TURP than during the simulated TURP. This could have been due to (1) the increased complexity of the operating room (OR) environment (2) the decreased quality of the image of the urethra and associated anatomy (compared with the VR simulator), or (3) the impairment of visual attentional control due to the increased levels of stress likely experienced in the OR. CONCLUSIONS: The findings suggest that the complexity of the environment surrounding VR simulators needs to be considered in the design of effective simulated training curricula. The study also provides support for the use of eye-tracking technology to assess the content validity of simulation and to examine psychomotor processes during live operations.

Jaatha: a fast composite-likelihood approach to estimate demographic parameters.

While information about a species' demography is interesting in its own right, it is an absolute necessity for certain types of population genetic analyses. The most widely used methods to infer a species' demographic history do not take intralocus recombination or recent divergence into account, and some methods take several weeks to converge. Here, we present Jaatha, a new composite-likelihood method that does incorporate recent divergence and is also applicable when intralocus recombination rates are high. This new method estimates four demographic parameters. The accuracy of Jaatha is comparable to that of other currently available methods, although it is superior under certain conditions, especially when divergence is very recent. As a proof of concept, we apply this new method to estimate demographic parameters for two closely related wild tomato species, Solanum chilense and S. peruvianum. Our results indicate that these species likely diverged 1.44·N generations ago, where N is the effective population size of S. chilense, and that some introgression between these species continued after the divergence process initiated. Furthermore, S. peruvianum likely experienced a population expansion following speciation.

Control theoretic interpretation of directional motion preferences in optic flow processing interneurons.

In this article, we formalize the processing of optic flow in identified fly lobula plate tangential cells and develop a control theoretic framework that suggests how the signals of these cells may be combined and used to achieve reflex-like navigation behavior. We show that this feedback gain synthesis task can be cast as a combined static state estimation and linear feedback control problem. Our framework allows us to analyze and determine the relationship between optic flow measurements and actuator commands, which greatly simplifies the implementation of biologically inspired control architectures on terrestrial and aerial robotic platforms.

A nonlinear dynamics model for simulating long range correlations of cognitive bistability.

Simulation results of bistable perception due to ambiguous visual stimuli are presented which are obtained with a behavioral nonlinear dynamics model using perception-attention-memory coupling. This model provides an explanation of recent experimental results of Gao et al. (Cogn Process 7:105-112, 2006a) and it supports their speculation that the fractal character of perceptual dominance time series may be understood in terms of nonlinear and reentrant dynamics of brain processing. Percept reversals are induced by attention fatigue and noise, with an attention bias which balances the relative percept duration. Dynamical coupling of the attention bias to the perception state introduces memory effects leading to significant long range correlations of perceptual duration times as quantified by the Hurst parameter H > 0.5 (Mandelbrot, The fractal geometry of nature, 1991), in agreement with Gao et al. (Cogn Process 7:105-112, 2006a).

Assessment of surgical competence at carotid endarterectomy under local anaesthesia in a simulated operating theatre.

BACKGROUND: Methods of surgical training that do not put patients at risk are desirable. A high-fidelity simulation of carotid endarterectomy under local anaesthesia was tested as a tool for assessment of vascular surgical competence, as an adjunct to training. METHODS: Sixty procedures were performed by 30 vascular surgeons (ten junior trainees, ten senior trainees and ten consultants) in a simulated operating theatre. Each performed in a non-crisis scenario followed by a crisis scenario. Performance was assessed live by means of rating scales for technical and non-technical skills. RESULTS: There was a significant difference in technical skills with ascending grade for both generic and procedure-specific technical skill scores in both scenarios (P < 0.001 for all comparisons). Similarly, there was also a significant difference in non-technical skill with ascending grade for both scenarios (P < 0.001). There was a highly significant correlation between technical and non-technical performance in both scenarios (non-crisis: r(s) = 0.80, P < 0.001; crisis: r(s) = 0.85, P < 0.001). Inter-rater reliability was high (alpha > or = 0.80 for all scales). CONCLUSION: High-fidelity simulation offers competency-based assessment for all grades and may provide a useful training environment for junior trainees and more experienced surgeons.

Simulation analysis of ASM/Takács models in the BSM1 configuration.

The present paper summarizes the outlines of a simulation analysis study conducted on the BSM1 (formerly the COST benchmark) configuration using activated sludge models (ASM1, ASM2d and ASM3) coupled to Takács settler model. The prime objective was to develop reliable simulation software programs to implement these complex models according to the working conditions of a realistic plant. The analysis focused on comparing the steady state predictions of models ASM1/ ASM3 when imposed to pulse/ step type disturbances on wastewater characteristics and control variables, then on assessing the capability of the simulated configuration for bio-P removal using model ASM2d. Section 1 of the paper briefly presents problem definition/ solution approach while section 2 demonstrates some examples showing the main indications of the simulation analysis. ASM1/ ASM3 predictions indicate the presence of some significant differences between both models that could be related to their underlying concepts. ASM2d simulations show that adverse effects on the permissible limit of effluent's ammonia concentration should be expected when the plant is operated to achieve dual nutrient removal.

In vitro screening of drug metabolism during drug development: can we trust the predictions?

Due to the importance of drug metabolism for drug action, side effects, interactions and interindividual differences, in vitro assays of drug metabolism are widely employed during drug development. Validation of the in vitro systems is still in a rather elementary stage, but some tentative conclusions about their performance for predicting various important in vivo characteristics of drugs can be attempted. So far, prediction of drug-drug interactions on the basis of in vitro screens has been advanced to a relatively reliable level. Prediction of other important characteristics, such as metabolic stability and consequent half-life and dosage schedule, or induction potential, is less reliable. Although metabolite profiles and participating enzymes are predicted quite accurately with advanced analytical techniques, an important problem here is the lack of reliable methods to detect reactive metabolites or products, which may be of significance regarding toxicity.

A beef herd model for simulating feed intake, animal performance, and manure excretion in farm systems.

A beef herd submodel was created for integration with other farm components to form a whole-farm model capable of simulating a wide range of beef production systems. This herd submodel determined the best available feed or feed mix to meet the fiber, energy, and protein requirements for each of up to six animal groups on the farm. The groups comprised any combination of cows, nursing calves, young heifers, yearling heifers, stockers, and finishing cattle. Protein, energy, and mineral requirements were determined for each group using the Cornell Net Carbohydrate and Protein System, Level 1. Diets were formulated to meet these requirements with available feeds, and the resulting feed intake, growth, and manure DM and nutrient (N, P, and K) excretions were predicted. Required feed characteristics included CP, ruminally degradable protein, acid detergent insoluble protein, NDF, P, and K concentrations. Feed intake was predicted by considering energy intake, potentially limited by fill, and exceeding a minimum roughage requirement. Fill and roughage limits were functions of feed NDF concentrations adjusted to consider particle size distribution and the relative rate of ruminal digestibility or the physical effectiveness of the fiber. The herd submodel was verified to predict feed intakes, nutrient requirements, diets, and manure excretions similar to those recommended or measured for beef animals. Incorporation of the beef herd submodel with other farm components, including crop growth (alfalfa, grass, corn, small grain, and soybean), harvest, storage, feeding, grazing, and manure handling, provided the Integrated Farm System Model. This comprehensive farm-simulation model is a useful research and teaching tool for evaluating and comparing the long-term performance, economics, and environmental impact of beef, dairy, and crop production systems.

Testing the raters: inter-rater reliability of standardized anaesthesia simulator performance.

PURPOSE: Assessment of physician performance has been a subjective process. An anaesthesia simulator could be used for a more structured and standardized evaluation but its reliability for this purpose is not known. We sought to determine if observers witnessing the same event in an anaesthesia simulator would agree on their rating of anaesthetist performance. METHODS: The study had the approval of the research ethics board. Two one-hour clinical scenarios were developed, each containing five anaesthetic problems. For each problem, a rating scale defined the appropriate score (no response to the situation: score = 0; compensating intervention defined as physiological correction: score = 1; corrective treatment: defined as definitive therapy score = 2). Video tape recordings, for assessment of inter-rater reliability, were generated through role-playing with recording of the two scenarios three times each resulting in a total of 30 events to be evaluated. Two clinical anaesthetists, uninvolved in the development of the study and the clinical scenarios, reviewed and scored each of the 30 problems independently. The scores produced by the two observers were compared using the kappa statistic of agreement. RESULTS: The raters were in complete agreement on 29 of the 30 items. There was excellent inter-rater reliability (= 0.96, P < 0.001). CONCLUSIONS: The use of videotapes allowed the scenarios to be scored by reproducing the same event for each observer. There was excellent inter-rater agreement within the confines of the study. Rating of video recordings of anaesthetist performance in a simulation setting can be used for scoring of performance. The validity of the scenarios and the scoring system for assessing clinician performance have yet to be determined.