Journal of Chemometrics

Cover image for Journal of Chemometrics

January 2010

Volume 24, Issue 1

Pages 1–42

  1. Research Articles

    1. Top of page
    2. Research Articles
    1. Modeling and predicting binding affinity of phencyclidine-like compounds using machine learning methods (pages 1–13)

      Ozlem Erdas, Erdem Buyukbingol, Ferda Nur Alpaslan and Adeboye Adejare

      Article first published online: 23 DEC 2009 | DOI: 10.1002/cem.1265

      In this study, molecular electrostatic potential surfaces of phencyclidine-like compounds are modeled and visualized in order to extract features that are useful in predicting binding affinities. In modeling, the Cartesian coordinates of MEP surface points are mapped onto a spherical self-organizing map. The resulting maps are visualized using electrostatic potential values. These values also provide features for a prediction system. Support vector machines and partial least-squares method are used for predicting binding affinities of compounds.

    2. A diagonal measure and a local distance matrix to display relations between objects and variables (pages 14–21)

      Gergely Tóth and Pál Szepesváry

      Article first published online: 22 DEC 2009 | DOI: 10.1002/cem.1267

      We introduced two concepts to develop a method, where a proper change of rows and columns of a data matrix yields patterns of significant new information. By using our diagonal matrix measure and local distance matrix the similar objects were arranged close to each other in the data matrix and simultaneously the variables responsible for their similarity were collected close to the diagonal part defined by these objects.

    3. On further application of rmath image as a metric for validation of QSAR models (pages 22–33)

      Indrani Mitra, Partha Pratim Roy, Supratik Kar, Probir Kumar Ojha and Kunal Roy

      Article first published online: 8 DEC 2009 | DOI: 10.1002/cem.1268

      The present paper attempts to show that “true rmath image(LOO)” statistic calculated based on a QSAR model derived from a undivided data set with application of variable selection strategy at each cycle of leave-one-out validation may reflect external validation characteristics of the developed model thus obviating the requirement of splitting of the data set into training and test sets. This approach may be helpful in developing QSAR models from small data sets.

    4. Optimal designs for estimating the parameters in weighted power-mean-mixture models (pages 34–42)

      R. L. J. Coetzer and W. W. Focke

      Article first published online: 23 DEC 2009 | DOI: 10.1002/cem.1271

      In the mixing of fluids, a mixture may be viewed conceptually as a hypothetical collection of fluid clusters. A particular flexible form is obtained from using generalized weighted-power-means with the weighting based on global mole fractions. In this paper, we present optimal designs for estimating the parameters in the generalized weighted-power-mean mixture models, which may be nonlinear in the pure and binary interaction parameters. We illustrate the practical value of applying optimal designs for mixture variables through design efficiencies.