MPGraph: multi‐view penalised graph clustering for predicting drug–target interactions

Identifying drug–target interactions has been a key step for drug repositioning, drug discovery and drug design. Since it is expensive to determine the interactions experimentally, computational methods are needed for predicting interactions. In this work, the authors first propose a single‐view penalised graph (SPGraph) clustering approach to integrate drug structure and protein sequence data in a structural view. The SPGraph model does clustering on drugs and targets simultaneously such that the known drug–target interactions are best preserved in the clustering results. They then apply the SPGraph to a chemical view with drug response data and gene expression data in NCI‐60 cell lines. They further generalise the SPGraph to a multi‐view penalised graph (MPGraph) version, which can integrate the structural view and chemical view of the data. In the authors'' experiments, they compare their approach with some comparison partners, and the results show that the SPGraph could improve the prediction accuracy in a small scale, and the MPGraph can achieve around 10% improvements for the prediction accuracy. They finally give some new targets for 22 Food and Drug Administration approved drugs for drug repositioning, and some can be supported by other references.Inspec keywords: graphs, drug delivery systems, drugs, proteins, molecular biophysics, molecular configurations, optimisation, eigenvalues and eigenfunctions, Laplace equations, cancer, cellular biophysics, gene therapy, medical computingOther keywords: MPGraph, multiview penalised graph clustering, drug‐target interactions, drug repositioning, drug discovery, drug design, computational methods, single‐view penalized graph clustering approach, drug structure, protein sequence data, SPGraph model, optimisation problem, spectral clustering, eigenvalue decomposition, Laplacian model, gene expression data, NCI‐60 cell lines     

Allocation of variance reduction targets under the influence of supplier interaction

A common quality improvement strategy used by manufacturers is to periodically allocate quality improvement targets among their suppliers. We propose a formal modelling and optimization approach for assessing quality improvement targets for suppliers. In this approach it is understood that a manufacturer's quality improvement results from reductions in supplier process variances, which occurs only through investments in learning. A constrained nonlinear optimization model is developed for determining an optimal allocation of variance reduction target that minimizes expected total cost, where the relationship between performance measures and the set of design parameters is generally represented by second-order polynomial functions. An example in the fabrication of a tyre tread compound is used both to demonstrate the implementation of our proposed models as well as to provide an empirical comparison of optimal learning rates for different functional relationships between the performance measures and the set of design parameters.     

A probabilistic approach to identify putative drug targets in biochemical networks

Network-based drug design holds great promise in clinical research as a way to overcome the limitations of traditional approaches in the development of drugs with high efficacy and low toxicity. This novel strategy aims to study how a biochemical network as a whole, rather than its individual components, responds to specific perturbations in different physiological conditions. Proteins exerting little control over normal cells and larger control over altered cells may be considered as good candidates for drug targets. The application of network-based drug design would greatly benefit from using an explicit computational model describing the dynamics of the system under investigation. However, creating a fully characterized kinetic model is not an easy task, even for relatively small networks, as it is still significantly hampered by the lack of data about kinetic mechanisms and parameters values. Here, we propose a Monte Carlo approach to identify the differences between flux control profiles of a metabolic network in different physiological states, when information about the kinetics of the system is partially or totally missing. Based on experimentally accessible information on metabolic phenotypes, we develop a novel method to determine probabilistic differences in the flux control coefficients between the two observable phenotypes. Knowledge of how differences in flux control are distributed among the different enzymatic steps is exploited to identify points of fragility in one of the phenotypes. Using a prototypical cancerous phenotype as an example, we demonstrate how our approach can assist researchers in developing compounds with high efficacy and low toxicity.     

Multiobjective optimization by genetic algorithms: application to safety systems

When attempting to optimize the design of engineered systems, the analyst is frequently faced with the demand of achieving several targets (e.g. low costs, high revenues, high reliability, low accident risks), some of which may very well be in conflict. At the same time, several requirements (e.g. maximum allowable weight, volume etc.) should also be satisfied. This kind of problem is usually tackled by focusing the optimization on a single objective which may be a weighed combination of some of the targets of the design problem and imposing some constraints to satisfy the other targets and requirements. This approach, however, introduces a strong arbitrariness in the definition of the weights and constraints levels and a criticizable homogenization of physically different targets, usually all translated in monetary terms.The purpose of this paper is to present an approach to optimization in which every target is considered as a separate objective to be optimized. For an efficient search through the solution space we use a multiobjective genetic algorithm which allows us to identify a set of Pareto optimal solutions providing the decision maker with the complete spectrum of optimal solutions with respect to the various targets. Based on this information, the decision maker can select the best compromise among these objectives, without a priori introducing arbitrary weights.     

Statistical design of position-encoded microsphere arrays

We propose a microsphere array device with microspheres having controllable positions for error-free target identification. We conduct a statistical design analysis to select the optimal distance between the microspheres as well as the optimal temperature. Our design simplifies the imaging and ensures a desired statistical performance for a given sensor cost. Specifically, we compute the posterior Cramér-Rao bound on the errors in estimating the unknown target concentrations. We use this performance bound to compute the optimal design variables. We discuss both uniform and sparse concentration levels of targets, and replace the unknown imaging parameters with their maximum likelihood estimates. We illustrate our design concept using numerical examples. The proposed microarray has high sensitivity, efficient packing, and guaranteed imaging performance. It simplifies the imaging analysis significantly by identifying targets based on the known positions of the microspheres. Potential applications include molecular recognition, specificity of targeting molecules, protein-protein dimerization, high throughput screening assays for enzyme inhibitors, drug discovery, and gene sequencing.     

Hydrophilic interaction chromatography-electrospray mass spectrometry analysis of polar compounds for natural product drug discovery

For the drug discovery efforts currently taking place within the pharmaceutical industry, natural product extracts have been found to provide a valuable source of molecular diversity which is complementary to that provided by traditional synthetic organic methods or combinatorial chemistry. However, there exists a need for analytical tools that can facilitate the separation and characterization of components from these sources in a rapid manner. Specifically, the evaluation of highly polar compounds (i.e., compounds that cannot be retained on traditional reversed-phase stationary phases) has been challenging, and a hydrophilic interaction chromatography-electrospray ionization mass spectrometry (HILIC-ESI-MS) method was developed to meet this need. In this investigation, amide-, Polyhydroxyethyl Aspartamide-, and cyclodextrin-based packings provided superior performance for the analysis of a set of polar natural product compounds. The properties of the mobile-phase buffers also greatly impacted the separations, and relative to other volatile buffering agents, ammonium acetate at a concentration of approximately 6.5 mM was determined to facilitate optimal HILIC retention, reproducibility, and durability. An optimized HILIC-ESI-MS system was successfully applied for the analysis of complex natural product mixtures. The techniques described in this report should also prove useful for the analysis of polar compounds from synthetic sources of molecular diversity such as combinatorial chemistry.     

Optimizing Steering Angle Predictive Convolutional Neural Network for Autonomous Car

Deep learning techniques, particularly convolutional neural networks (CNNs), have exhibited remarkable performance in solving vision-related problems, especially in unpredictable, dynamic, and challenging environments. In autonomous vehicles, imitation-learning-based steering angle prediction is viable due to the visual imagery comprehension of CNNs. In this regard, globally, researchers are currently focusing on the architectural design and optimization of the hyperparameters of CNNs to achieve the best results. Literature has proven the superiority of metaheuristic algorithms over the manual-tuning of CNNs. However, to the best of our knowledge, these techniques are yet to be applied to address the problem of imitation-learning-based steering angle prediction. Thus, in this study, we examine the application of the bat algorithm and particle swarm optimization algorithm for the optimization of the CNN model and its hyperparameters, which are employed to solve the steering angle prediction problem. To validate the performance of each hyperparameters’ set and architectural parameters’ set, we utilized the Udacity steering angle dataset and obtained the best results at the following hyperparameter set: optimizer, Adagrad; learning rate, 0.0052; and nonlinear activation function, exponential linear unit. As per our findings, we determined that the deep learning models show better results but require more training epochs and time as compared to shallower ones. Results show the superiority of our approach in optimizing CNNs through metaheuristic algorithms as compared with the manual-tuning approach. Infield testing was also performed using the model trained with the optimal architecture, which we developed using our approach.     

Microgenetic multiobjective reconfiguration algorithm considering power losses and reliability indices for medium voltage distribution network

This study proposes and applies an evolutionary-based approach for multiobjective reconfiguration in electrical power distribution networks. In this model, two types of indicators of power quality are minimised: (i) power system's losses and (ii) reliability indices. Four types of reliability indices are considered. A microgenetic algorithm ('GA) is used to handle the reconfiguration problem as a multiobjective optimisation problem with competing and non-commensurable objectives. In this context, experiments have been conducted on two standard test systems and a real network. Such problems characterise typical distribution systems taking into consideration several factors associated with the practical operation of medium voltage electrical power networks. The results show the ability of the proposed approach to generate well-distributed Pareto optimal solutions to the multiobjective reconfiguration problem. In the systems adopted for assessment purposes, our proposed approach was able to find the entire Pareto front. Furthermore, better performance indexes were found in comparison to the Pareto envelope-based selection algorithm 2 (PESA 2) technique, which is another well-known multiobjective evolutionary algorithm available in the specialised literature. From a practical point of view, the results established, in general, that a compact trade-off region exists between the power losses and the reliability indices. This means that the proposed approach can recommend to the decision maker a small set of possible solutions in order to select from them the most suitable radial topology.     

Specific affinity extraction method for small molecule-binding proteins

Validation of the targets of candidate drugs is critical for rapid and efficient drug discovery and development and for understanding the pharmacological action and potential toxicities of the prospective therapeutic agent. Due to the nonspecific binding of abundant proteins to small molecule-immobilized gels, it is difficult to identify the protein targets of small molecules from crude biological samples by affinity extraction. To address this problem, we have developed an affinity gel for the specific extraction of small molecule-binding proteins. We immobilized small molecules on the agarose gel through a disulfide linker that is cleavable by mild reduction. This system has allowed specific and noncovalent complex formation between the small molecule and the target protein, keeping the effect of the nonspecific abundant proteins adsorbed on both the linker and gel surface to minimum. By preparing this affinity matrix with deoxycholate as a model small molecule, we captured two independent deoxycholate-binding proteins of different affinities from mouse ascites, anti-deoxycholate antibody, and serum albumin. As other proteins were not captured, this affinity extraction method should contribute significantly to the accurate and rapid drug discovery and development.     

DEA cross-efficiency evaluation based on satisfaction degree: an application to technology selection

Data envelopment analysis (DEA) has been extended to cross-efficiency evaluation to provide better discrimination and ranking of decision-making units (DMUs). However, the non-uniqueness of optimal weights in the traditional DEA models (CCR and BCC models) has reduced the usefulness of the DEA cross-efficiency evaluation method. To solve this problem, we introduce the concept of the satisfaction degree of a DMU towards a set of optimal weights for another DMU. Then, a new DEA cross-efficiency evaluation approach, which contains a maxmin model and two algorithms, is proposed based on the satisfaction degrees of the DMUs. Our maxmin model and algorithm 1 can obtain for each DMU an optimal set of weights that maximises the least satisfaction degrees among all the other DMUs. Further, our algorithm 2 can then be used to guarantee the uniqueness of the optimal weights for each DMU. Finally, our approach is applied to a real-world case study of technology selection.     

Optimal design of capacitated production networks

In this work we present a mixed-integer model for the optimal design of production/transportation systems. In contrast to standard design problems, our model is originally based on a coupled system of differential equations capturing the dynamics of manufacturing processes and stocks. The problem is to select an optimal parameter configuration from a predefined set such that respective constraints are fulfilled. We focus on single commodity flows over large time scales as well as highly interconnected networks and propose a suitable start heuristic to ensure feasibility and to speed up the solution procedure.     

Statistical total correlation spectroscopy scaling for enhancement of metabolic information recovery in biological NMR spectra

The high level of complexity in nuclear magnetic resonance (NMR) metabolic spectroscopic data sets has fueled the development of experimental and mathematical techniques that enhance latent biomarker recovery and improve model interpretability. We previously showed that statistical total correlation spectroscopy (STOCSY) can be used to edit NMR spectra to remove drug metabolite signatures that obscure metabolic variation of diagnostic interest. Here, we extend this "STOCSY editing" concept to a generalized scaling procedure for NMR data that enhances recovery of latent biochemical information and improves biological classification and interpretation. We call this new procedure STOCSY-scaling (STOCSY(S)). STOCSY(S) exploits the fixed proportionality in a set of NMR spectra between resonances from the same molecule to suppress or enhance features correlated with a resonance of interest. We demonstrate this new approach using two exemplar data sets: (a) a streptozotocin rat model (n = 30) of type 1 diabetes and (b) a human epidemiological study utilizing plasma NMR spectra of patients with metabolic syndrome (n = 67). In both cases significant biomarker discovery improvement was observed by using STOCSY(S): the approach successfully suppressed interfering NMR signals from glucose and lactate that otherwise dominate the variation in the streptozotocin study, which then allowed recovery of biomarkers such as glycine, which were otherwise obscured. In the metabolic syndrome study, we used STOCSY(S) to enhance variation from the high-density lipoprotein cholesterol peak, improving the prediction of individuals with metabolic syndrome from controls in orthogonal projections to latent structures discriminant analysis models and facilitating the biological interpretation of the results. Thus, STOCSY(S) is a versatile technique that is applicable in any situation in which variation, either biological or otherwise, dominates a data set at the expense of more interesting or important features. This approach is generally appropriate for many types of NMR-based complex mixture analyses and hence for wider applications in bioanalytical science.     

A new strategy for assessing sensitivities in biochemical models

An integral part of any systems biology approach is the modelling and simulation of the respective system under investigation. However, the values of many parameters of the system have often not been determined or are not identifiable due to technical experimental difficulties or other constraints. Sensitivity analysis is often employed to quantify the importance of each of the model's parameters in the behaviour of the system. This approach can also be useful in identifying those parts of the system that are most sensitive with the potential of becoming drug targets. A problem of the commonly used methods of sensitivity analysis is that they constitute local methods meaning that they depend directly on the exact parameter space, which in turn is not known exactly. One way to circumvent this problem is to carry out sensitivity analysis over a wide range of values for all parameters, but this is handicapped by expensive computations when the systems are high dimensional. Another approach is to employ global sensitivity analysis, which in this context is mostly based on random sampling methods. In this paper we present an efficient approach that involves using numerical optimizing methods that search a wide region of parameter space for a given model to determine the maximum and minimum values of its metabolic control coefficients. A relevant example for drug development is presented to demonstrate the strategy using the software COPASI.     

Liposome and lipid bilayer arrays towards biosensing applications

Sensitive and selective biosensors for high-throughput screening are having an increasing impact in modern medical care. The establishment of robust protein biosensing platforms however remains challenging, especially when membrane proteins are involved. Although this type of proteins is of enormous relevance since they are considered in >60% of the pharmaceutical drug targets, their fragile nature (i.e., the requirement to preserve their natural lipid environment to avoid denaturation and loss of function) puts strong additional prerequisites onto a successful biochip. In this review, the leading approaches to create lipid membrane-based arrays towards the creation of membrane protein biosensing platforms are described. Liposomes assembled in micro- and nanoarrays and the successful set-ups containing functional membrane proteins, as well as the use of liposomes in networks, are discussed in the first part. Then, the complementary approaches to create cell-mimicking supported membrane patches on a substrate in an array format will be addressed. Finally, the progress in assembling free-standing (functional) lipid bilayers over nanopore arrays for ion channel sensing will be reported. This review illustrates the rapid pace by which advances are being made towards the creation of a heterogeneous biochip for the high-throughput screening of membrane proteins for diagnostics, drug screening, or drug discovery purposes.     

An integrated model for designing and optimising an international logistics network

Supply-chain configuration has recently gained increasing attention both from the practitioner's perspective and as a research area. This paper proposes an integrated model for designing and optimising international logistics networks. It consists of a mixed integer linear programming model and a data-mapping section (i.e. methodological guidelines for gathering and processing the data necessary to set up the model). It has been specifically developed for solving the configuration problem for supply chains characterised by a complexity level typical of real-life global logistics networks. Although this topic is well understood and well elaborated at a technical level in the extant literature, it still presents obstacles in practice especially in terms of dealing with real-life complexity, service-level constraints and data mapping. Thus, we developed our integrated approach with the aim to fill these gaps. We designed our model for dealing with multiple-layer, single location-layer, multiple-commodity and time-constrained logistics networks, to be implemented in a single period time horizon and in a deterministic environment. The proposed approach represents an innovative contribution to the existing body of scientific knowledge and facilitates the data gathering and processing activities, which are largely recognised as complex and time-consuming processes for the management of logistics activities.     

Identification of molecular target of AMP-activated protein kinase activator by affinity purification and mass spectrometry

We show an efficient method to identify molecular targets of small molecular compounds by affinity purification and mass spectrometry. Binding proteins were isolated from target cell lysate using affinity columns, which immobilized the active and inactive compounds. All proteins bound to these affinity columns were eluted by digestion using trypsin and then were identified by mass spectrometry. The specific binding proteins to the active compound, a candidate for molecular targets, were determined by subtracting the identified proteins in an inactive compound-immobilized affinity column from that in an active compound-immobilized affinity column. This method was applied to identification of molecular targets of D942, a furancarboxylic acid derivative, which increases glucose uptake in L6 myocytes through AMP-activated protein kinase (AMPK) activation. To elucidate the mechanism of AMPK activation by D942, affinity columns that immobilized D942 and its inactive derivative, D768, were prepared, and the binding proteins were purified from L6 cell lysate. NAD(P)H dehydrogenase [quinone] 1 (complex I), which was shown as one of the specific binding proteins to D942 by subtracting the binding proteins to D768, was partially inhibited by D942, not D768. Because inhibition of complex I activity led to a decrease in the ATP/AMP ratio, and the change in the ATP/AMP ratio triggered AMPK activation, we identified complex I as a potential protein target of AMPK activation by D942. This result shows our approach can provide crucial information about the molecular targets of small molecular compounds, especially target proteins not yet identified.     

Feasibility of supercritical fluid chromatography/mass spectrometry of polypeptides with up to 40-mers

Supercritical fluid chromatography (SFC) provides a number of advantages over traditional HPLC such as speed, practical use of longer columns, a normal-phase retention mechanism, and reduced use of organic solvents. Yet, it has been a technique traditionally limited to relatively nonpolar compounds. The nature of SFC mobile and stationary phases did not allow the elution of ionic compounds or of peptides, except, in the latter case, for the most hydrophobic peptides. The characterization of peptides is critically important for drug discovery and development in the pharmaceutical industry, as well as for a variety of other important applications. Here, for the first time to our knowledge, we show that relatively large peptides (at least 40 mers), containing a variety of acidic and basic residues, can be eluted in SFC. We used trifluoroacetic acid as additive in a CO2/methanol mobile phase to suppress deprotonation of peptide carboxylic acid groups and to protonate peptide amino groups. A 2-ethylpyridine bonded silica column, which was specifically developed for SFC, was used for the majority of this work. The relatively simple mobile phase was compatible with mass spectrometric detection.