Separating drugs from nondrugs: a statistical approach using atom pair distributions

A computational approach to quantify the druglike character of chemical compounds is presented. For this purpose, the distribution of atom types and their pair-wise combinations in known drugs and nondrugs was examined. Statistical analysis of the occurrence probabilities was used to derive a drug-likeliness score on a logarithmic scale. "Typical" pharmaceutical agents exhibit scores greater than 0.3, while for ordinary substances, values below 0 are expected. Although any kind of fitting or error minimization scheme is absent in this method, confirmed drugs are predicted with an accuracy of at least 71%. Many falsely predicted nondrugs were found to closely resemble actual drugs or to contain unsuitable substitution patterns that can easily be ruled out by applying medicinal knowledge. As the outlined method is computationally inexpensive, this drug-likeliness score can therefore be used as a filter for the in silico screening of large substance databases.     

Radiative recombination rates with different statistical distributions

The isotope shift and hyperfine structure in a rhenium hollow cathode discharge was studied for transitions of the type 5d ⁵6s7s 5d ⁵6s6p and 5d ⁵6s6d 5d ⁵6s6p through Doppler-free saturation absorption laserspectroscopy and high resolution interferometry. Taking configuration mixing in the lower levels of 5d ⁵6s6p under consideration, we obtain average configuration isotope shift values for 5d ⁵6s7s of –1760(100) MHz and for 5d ⁵6s6d of –1970(200) MHz. These experimental values compare extremely well with the theoretically predicted configuration isotope shifts in rhenium, based on pseudo-relativistic Hartree-Fock calculations, of –1710 MHz and –1940 MHz, resp. In addition hyperfine structure constants for rhenium levels of 5d ⁵6s6d are reported here for the first time.Research scientist from the University of Istanbul, Turkey     

Effects of polarization on charge distributions and integral approximations

Polarization factors derived in a previous paper [1] are presented for 1s, 2s, 2p and 2p orbitals on two centers. The formulas for gross atomic populations and integral approximations are tested in diatomics. It is found that our definition of atomic charges proposed earlier is far superior to the Mulliken approximation. With proper partitioning into charge and hybrid parts the total dipole moment of the wavefunction can be obtained with great accuracy. The application of the formulas to two-center hybrid repulsion integrals is less satisfactory in the lowest-order approximation.

---

Zusammenfassung Die Polarisationsfaktoren einer früheren Arbeit [1] werden für 1s, 2s-, 2p- und 2p-Orbitale an zwei Zentren angegeben. Die Formeln für Atomladungsverteilungen und Integralapproximationen werden in zweiatomigen Molekülen geprüft. Man findet, da\ unsere vorgeschlagene Definition von Atomladungen der Mulliken-NÄherung weit überlegen ist. Bei geeigneter Aufteilung in Ladungs- und Hybridanteil kann das Dipolmoment der Wellenfunktion mit gro\er Genauigkeit erhalten werden. Die Anwendung der Formeln auf Zwei-Zentren-Hybridintegrale der Elektronenwechselwirkung ist weniger zufriedenstellend in der niedrigsten Approximationsstufe.

     

Rate of radiated power loss with different statistical distributions

We compute the energy emitted in the radiative recombination process (or radiated power loss) of point particles of chargeZe and ?Z′e with nonrelativistic velocities and reduced massM, useful in a great variety of problems, averaged over Maxwellian and non-Maxwellian distributions. This has been accomplished using a recent parametrization, obtained by us, of the nonrelativistic, radiative recombination cross section in the dipole approximation, valid for all values of the energy.     

Phase diagram of complex fluids using an efficient integral equation method

We present an adaptive technique for the determination of the phase diagram of fluids within the integral equation theory. It enables an efficient and accurate systematic mapping of the thermodynamic space in order to construct the binodal and spinodal lines. Results are obtained with the thermodynamically consistent integral equation proposed by Sarkisov [J. Chem. Phys. 114, 9496 (2001)] within the tangent linear technique that yields an exact differentiation of correlation functions. The generality of the numerical approach is assessed by determining both the liquid-vapor coexistence and the critical parameters of the generalized Lennard-Jones (n,6) potentials with varying repulsive part, including the hard-sphere limit.     

Prior statistical distributions implied by dynamical threshold laws for molecular collisions

Dynamical threshold laws imply a particular translational energy dependence of the prior distribution function used in surprisal theory. The microcanonical prior distribution is not compatible with collinear and coplanar threshold laws but is compatible with three-dimensional dynamics.     

Selective ion binding by protein probed with the statistical mechanical integral equation theory

Selective ion binding by human lysozyme and its mutants is probed with the three-dimensional interaction site model theory which is the statistical mechanical integral equation theory. Preliminary and partial results of the study have been already published (Yoshida, N. et al. J. Am. Chem. Soc. 2006, 128, 12042-12043). The calculation was carried out for aqueous solutions of three different electrolytes, CaCl2, NaCl, and KCl, and for four different mutants of the human lysozyme: wild type, Q86D, A92D, and Q86D/A92D, which have been studied experimentally. The discussion of this article focuses on the cleft that consists of amino acid residues from Q86 to A92. For the wild type of protein in the aqueous solutions of all the electrolytes studied, there are no distributions observed for the ions inside the cleft. The Q86D mutant shows essentially the same behavior with that of the wild type. The A92D mutant shows strong binding ability to Na+ in the recognition site, which is in accord with the experimental results. There are two isomers of the Q86D/A92D mutant, e.g., apo-Q86D/A92D and holo-Q86D/A92D. Although both isomers exhibit the binding ability to the Na+ and Ca2+ ions, the holo isomer shows much greater affinity compared with the apo isomer. Regarding the selective ion binding of the holo-Q86D/A92D mutant, it shows greater affinity to Ca2+ than to Na+, which is also consistent with the experimental observation.     

Modeling mercury porosimetry using statistical mechanics

We consider mercury porosimetry from the perspective of the statistical thermodynamics of penetration of a nonwetting liquid into a porous material under an external pressure. We apply density functional theory to a lattice gas model of the system and use this to compute intrusion/extrusion curves. We focus on the specific example of a Vycor glass and show that essential features of mercury porosimetry experiments can be modeled in this way. The lattice model exhibits a symmetry that provides a direct relationship between intrusion/extrusion curves for a nonwetting fluid and adsorption/desorption isotherms for a wetting fluid. This relationship clarifies the status of methods that are used for transforming mercury intrusion/extrusion curves into gas adsorption/desorption isotherms. We also use Monte Carlo simulations to investigate the nature of the intrusion and extrusion processes.     

The numerical solution of fujita's integral equation to provide estimates of molecular weight distributions

A standard model of the behavior of polymers under ultracentrifugation results in Fujita's equation for their molecular weight distribution. Fujita's and related equations are examples of Fredholm integral equations of the first kind and are thus ill posed. Two methods are described for solving the equations numerically and hence providing estimates of the molecular weight distribution. The first method involves expanding the distribution in terms of orthogonal polynomials whose coefficients are calculated from estimates of the moments of the distribution. In the second method the distribution is reconstructed by using matrix singular-value decomposition techniques combined with an approximant expressed as a sum of B-splines. The potential and practical limitations associated with the methods are illustrated by numerical results from a series of tests on four problems designed to represent distributions with different modal properties.     

Chemical compound design using nuclear charge distributions

Finding optimal solutions to design problems in chemistry is hampered by the combinatorially large search space. We develop a general theoretical framework for finding chemical compounds with prescribed properties using nuclear charge distributions. The key is the reformulation of the design problem into an optimization problem on probability density functions in chemical space. In order to achieve tractability, a constrained search formalism on the nuclear charge distributions, which are non-negative, is used to reduce the dimensionality of the problem. Furthermore, we introduce approximations to the exact functional, as derived, for common design properties and constraints.     

Vector length distributions and polarizabilities of effective polymer chains: a statistical segment approach

The freely orienting model of a polymer chain is generalized by considering the distribution of vector lengths and polarizabilities of the statistical segments in the chain with a constant number of skeletal bonds in each of the segments The bonds in the segments are assumed to exist in their RIS (Rotational Isomeric States) conformations. The segment is characterized, i.e., its end-to-end length and polarizability distributions are computed. Bond polarizabilities, as determined by Denbigh, have been used for polyethylene and poly(cis-isoprene), and are assumed to be independent of the environment. Two methods are used to compute the chain length distribution from the length distribution of statistical segments: (i) an exact method, using a modified version of Chandrasekhar’s approach, originally formulated for chains of segments having constant length; and (ii) an alternative approach, which considers the series expansion of the Helmholtz Free Energy of an isolated chain, making the analysis computationally more viable without significant loss in accuracy. The averages of the chain end-to-end length distributions have been computed at 373 K for poly(cis-isoprene) and at 403 and 413 K for polyethylene. Also, chain polarizability is determined from the distribution of statistical segment lengths and polarizabilities. The results are in a form that can be used to obtain stress-deformation and optical anisotropy-deformation relationships of assemblies of chains, such as crosslinked networks.     

Similarity studies using statistical and genetical methods

This paper aims at demonstrating the applicability of statistical spectroscopy and genetic algorithms to the similarity studies. Statistical moments of the intensity distributions are used as a basis for defining similarity distances between pairs of model spectra. Model spectrum is taken as a sum of two Gaussian distributions characterized by different parameters. As a result, dissimilarity maps are presented.     

Computational comparison of theoretical and experimental molecular spectra using an integral transformation in solving analytical problems

The problem of the comparison of computed and experimental curves of the distribution of absorption and transmission coefficients or absorbances is considered on examples of infrared spectra. A technique based on the mathematical smoothing of the spectral curve by introducing an instrument function is applied. It is shown that this results in a significant increase in the correlation coefficient between the theoretical and experimental spectral curves because of the reduction of the effect of intrinsic errors of the theoretical models and of the variability of experimental conditions. This property can be used to develop procedures for the identification of individual compounds and mixture components and for their quantitative determination without using certified reference samples of composition and reference samples at all.     

Prediction of anion distributions using Pauling's second rule

Pauling's second crystal rule is shown to be able to predict in a general and simple way the distribution of anions in mixed oxyanion systems such as oxynitrides and oxyhalides with diverse compositions and structure types. Results are presented in a plot correlating the charge of the anions obtained from the observed occupancies in crystallographic positions with the calculated bond strength sums for these sites.     

Estimating probabilities of diabetes mellitus using neural networks

Classification problems are often encountered in medical diagnosis. This paper presents an introduction to classification theory and shows how artificial neural networks can be used for classification. We also map out a bootstrapped procedure for interval estimation of posterior probabilities. The entire procedure is illustrated using the diabetes mellitus data in Pima Indians.     

The impact of the resolution of the identity approximate integral method on modern ab initio algorithm development

The computation of the two-electron four-center integrals over gaussian basis functions is a significant component of the overall work of many ab initio methods used today. Improvements in the computational efficiency of the base algorithms have provided significant impact. Somewhat overlooked are methods that provide approximations to these integrals and their implementation in application software. A partial review of approximate integral techniques focused on the resolution of the identity (RI) four-center, two-electron integral approximation is given. The past and current uses of the RI algorithms are presented along with possibilities for further exploitation of the technology. Received: 14 January 1997 / Accepted: 11 March 1997     

Constructing a self-consistent integral baseline by using cubic splines

Various types of transformations require different baselines reflecting specificities of these transitions. The present work deals with the case when a degree of transformation is directly proportional to heat consumed or released. For such case, a baseline is named an integral baseline and is traditionally constructed by unnecessary simplifications. A new method is proposed as an alternative fast and robust computational method for baseline construction utilizing interpolating cubic splines. The method is self-consistent in the sense that it is free of needless assumptions and that it provides linearity between the degree of transformation and heat measured.     

Estimating protein-ligand binding affinity using high-throughput screening by NMR

Many of today's drug discovery programs use high-throughput screening methods that rely on quick evaluations of protein activity to rank potential chemical leads. By monitoring biologically relevant protein-ligand interactions, NMR can provide a means to validate these discovery leads and to optimize the drug discovery process. NMR-based screens typically use a change in chemical shift or line width to detect a protein-ligand interaction. However, the relatively low throughput of current NMR screens and their high demand on sample requirements generally makes it impractical to collect complete binding curves to measure the affinity for each compound in a large and diverse chemical library. As a result, NMR ligand screens are typically limited to identifying candidates that bind to a protein and do not give any estimate of the binding affinity. To address this issue, a methodology has been developed to rank binding affinities for ligands based on NMR screens that use 1D (1)H NMR line-broadening experiments. This method was demonstrated by using it to estimate the dissociation equilibrium constants for twelve ligands with the protein human serum albumin (HSA). The results were found to give good agreement with previous affinities that have been reported for these same ligands with HSA.     

New linear integral kinetic parameters assessment method based on an accurate approximate formula of temperature integral

This work concerns a proposition of a new assessment method to obtain kinetic parameters from nonisothermal solid-state kinetics, based on a new and accurate approximate formula of temperature integral. The new formula was derived numerically by a two-step linearly fitting process without using any further approximating series. The relative error involved in the activation energy has been estimated and found to be less than 0.001% in the practical range of 15 < x < 60. A comparison of the suggested approximations to published approximates has shown significant improvements in terms of accuracy at high and low x values. The validity of the new method has been confirmed by computing activation energy from experimental data. Moreover, two approaches have been proposed to determine the kinetic reaction model and preexponential factor based on the new approximate formula. The comparison of the obtained results arising from the application of the present method to others obtained by the most widely reported methods in the literature shows a remarkable preeminence of the new method.