Analyses of known protein–ligand interactions play an important role in designing novel and efficient drugs, contributing to drug discovery and development. Recently, machine learning methods have proven useful in the design of novel drugs, which utilize intelligent techniques to predict the outcome of unknown protein–ligand interactions by learning from the physical and geometrical properties of known protein–ligand interactions. The aim of this study is to work through a specific example of a novel computational method, namely compressed images for affinity prediction (CIFAP), in which binding affinities for structurally related ligands in complexes with human checkpoint kinase 1 (CHK1) are predicted. The CIFAP algorithm presented here relates published pIC 50 values of 57 compounds, derived from a thienopyridine pharmacophore, in complexes with CHK1 to their two-dimensional (2D) electrostatic potential images compressed in orthogonal dimensions. Patterns obtained from the 2D images are then used as inputs in regression and learning algorithms such as support vector regression (SVR) and adaptive neuro-fuzzy inference system (ANFIS) methods to validate the experimental pIC 50 values. This study revealed that the 2D image pixels in the vicinity of bound ligand surfaces provide more relevant information to make correlations with the empirical pIC 50 values. As compared with ANFIS, SVR gave rise to the lowest root mean square errors and the greatest correlations, suggesting that SVR could be a plausible choice of machine learning methods in predicting binding affinities by CIFAP. Copyright © 2013 John Wiley & Sons, Ltd.
A neural network model for predicting country-level concentrations of the fraction of particulates in the air with sizes less than 10 µm (PM10) has been developed using widely available sustainability and economical/industrial parameters as inputs. The model was trained and validated with the data for 23 European Union (EU) countries plus the EU27 as a group for the period from 2000 to 2008. The inputs for the model were selected using correlation analyses. Country-level PM10 concentration data that were used as a model output were obtained from the World Bank. The artificial neural network (ANN) model, created with inputs chosen by correlation analyses, has shown very good performance in the forecast of country-level PM10 concentrations. The mean absolute error for the ANN model prediction, in the case of most of the EU countries, was less than 13%, indicating stable and accurate predictions. The predictions obtained from the principal component regression model, which was trained and tested using the same datasets and input variables, had mean absolute errors from 20% to 150% for most of the countries. The wide availability of input parameters used in this model can overcome the problem of lack and scarcity of data in many countries, which can in turn prevent the determination of human exposure to PM10 at the national level. Copyright © 2013 John Wiley & Sons, Ltd.
]]>Héberger and colleagues [Trends Anal Chem 2010;29:101–109; J Chemometrics 2011;25:151–158] have introduced the sum of ranking differences as a measure for comparing models or methods and have demonstrated its applicability in a variety of settings. The sum of ranking differences is closely related to another distance measure for permutations, namely, the inversion number. In this note, we describe the inversion number along with some of its distributional properties and draw comparisons with the sum of ranking differences for model comparison. Copyright © 2013 John Wiley & Sons, Ltd.
]]>A grid potential analysis employing a novel approach of 3D quantitative structure–activity relationships (QSAR) as AutoGPA module in MOE2009.10 was performed on a dataset of 42 compounds of N-arylsulfonylindoles as anti-HIV-1 agents. The uniqueness of AutoGPA module is that it automatically builds the 3D-QSAR model on the pharmacophore-based molecular alignment. The AutoGPA-based 3D-QSAR model obtained in the present study gave the cross-validated Q2 value of 0.588, r2pred value of 0.701, r2m statistics of 0.732 and Fisher value of 94.264. The results of 3D-QSAR analysis indicated that hydrophobic groups at R1 and R2 positions and electron releasing groups at R3 position are favourable for good activity. To find similar analogues, virtual screening on ZINC database was carried out using generated AutoGPA-based 3D-QSAR model and showed good prediction. In addition to those mentioned earlier, in-silico ADME absorption, distribution, metabolism and excretion profiling and toxicity risk assessment test was performed, and results showed that majority of compounds from current dataset and newly virtually screened hits generated were within their standard limit. Copyright © 2013 John Wiley & Sons, Ltd.
]]>No abstract is available for this article.
]]>The development of a chromatographic procedure for an ultra-performance liquid chromatography can be a very time-consuming task, as the general approach for finding the appropriate operating conditions has been a trial-and-error process. The present study reports a novel approach in the field of ultra-performance liquid chromatography by using statistical experimental design based on Taguchi's method, which allows a complete separation of nine isocyanates present in a complex matrix. The resolution between two adjacent peaks was considered as a quality characteristic and transformed to a Taguchi signal-to-noise ratio. An orthogonal array L9 (34) was selected to analyze the effect of four chromatographic factors, that is, proportion of solvent, percent triethylamine (v/v), temperature (°C), and flow (mL min−1), with three levels each. The joint analysis performed to the significant factors achieved in the eight analyses of variance allowed to identify two methods to conduct a complete separation of all peaks. Six isocyanates were separated with the first method, with all factors at the lowest level. The remaining three isocyanates were separated with the second method, with the proportion of solvent at the highest level and the other factors at the lowest level. The overall Taguchi experimental design identified the proportion of solvent and the flow rate as major chromatographic factors. Finally, confirmatory experiments were performed with samples prepared with six and three isocyanates, confirming the complete separation of all isocyanates in the study.
The Taguchi methods provided a systematic and efficient methodology for this optimization, with considerably less effort than would be required for other optimizations techniques. Copyright © 2013 John Wiley & Sons, Ltd.
PARAFAC2 is applied in multiple research areas, for example, where data containing shifts are analysed, but it is a challenge to determine the appropriate number of components in the model. In this paper, it is hypothesized that the core consistency diagnostic, which is currently applied in, for example, PARAFAC1 can be used to determine model complexity in PARAFAC2. Theoretically, a PARAFAC1 model is fitted ‘inside’ the PARAFAC2 algorithm, and it should therefore be possible to apply the core consistency diagnostic from PARAFAC1 in PARAFAC2. To support this hypothesis, three different datasets, as well as simulated datasets, have been evaluated by means of PARAFAC2, and the core consistencies have been investigated. There is a general trend that if the core consistency is low, the model is overfitted as in PARAFAC1. Also, core consistency captures the true variation in the data, whereas small peaks are easily overlooked by visual inspection of noisy models. However, for determining the number of components in a PARAFAC2 model, we suggest usage of the core consistency in combination with other model parameters such as residuals, loadings, and split-half analysis. Copyright © 2013 John Wiley & Sons, Ltd.
]]>The multicomponent factorization of multivariate data often results in nonunique solutions. The so-called rotational ambiguity paraphrases the existence of multiple solutions that can be represented by the area of feasible solutions (AFS). The AFS is a bounded set that may consist of isolated subsets. The numerical computation of the AFS is well understood for two-component systems and is an expensive numerical process for three-component systems. In this paper, a new fast and accurate algorithm is suggested that is based on the inflation of polygons. Starting with an initial triangle located in a topologically connected subset of the AFS, an automatic extrusion algorithm is used to form a sequence of growing polygons that approximate the AFS from the interior. The polygon inflation algorithm can be generalized to systems with more than three components. The efficiency of this algorithm is demonstrated for a model problem including noise and a multicomponent chemical reaction system. Further, the method is compared with the recent triangle-boundary-enclosing scheme of Golshan, Abdollahi, and Maeder (Anal. Chem. 2011, 83, 836–841). Copyright © 2013 John Wiley & Sons, Ltd.
]]>The ICH Q2(R1) (International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use) guideline for testing linearity in validation of analytical procedures suggests that “linearity should be evaluated by visual inspection of a plot of signals as a function of analyte concentration or content.” The EP6-A guideline recommends more quantitative methods that compare straight-line and higher-order polynomial curve fits. In this paper, a new equivalence test is proposed to compare the quality of a straight-line fit to that of a higher-order polynomial. By using orthogonal polynomials and generalized pivotal quantity analysis, one may estimate the probability of equivalence between a straight line and a polynomial curve fit either in the assay signal space (the Y values) or in the concentration space (the X values). In the special case of the linear-to-quadratic polynomial comparison, an equivalence test may be constructed via a two one-sided T test. Copyright © 2013 John Wiley & Sons, Ltd.
]]>Orthogonal pre-processing (orthogonal projection) of spectral data is a common approach to generate analyte-specific information for use in multivariate calibration. The goal of this pre-processing is to remove from each spectrum the respective sample interferent contributions (spectral interferences from overlap, scatter, noise, etc.). Two approaches to accomplish orthogonal pre-processing are net analyte signal (NAS) and generalized least squares (GLS). Developed in this paper is the mathematical relationship between NAS and GLS. It is also realized that orthogonal NAS pre-processing can remove too much analyte signal and that the degree of interferent correction can be regulated. Similar to GLS, the degree of correction is accomplished by using a regularization (tuning) parameter to form generalized NAS (GNAS). Also developed in this paper is an alternative to GNAS and GLS based on generalized Tikhonov regularization (GTR). The mathematical relationships between GTR, GNAS, and GLS are derived. A result is the ability to express the model vector as the sum of two contributions: the orthogonal NAS contribution and a non-NAS contribution from the interferent components. Thus, rather than the usual situation of sequentially pre-processing data by either GNAS or GLS followed by model building with the pre-processed data, the methods of GTR, GNAS, and GLS are expressed as direct computations of model vectors allowing concurrent pre-processing and model building to occur. Simultaneous pre-processing and model forming are shown to be natural to the GTR process. Two near-infrared spectroscopic data sets are studied to compare the theoretical relationships between GTR, GNAS, and GLS. One data set covers basic calibration, and the other data set is for calibration maintenance. Filter factor representation is key to developing the interprocess relationships. Copyright © 2013 John Wiley & Sons, Ltd.
]]>