Search PubMed⌕ Search

Biomedical subjects

B Walczak

Publications and source records attributed to B Walczak.

13 recordsLinked to original sources
PubMed2002

Principal component analysis of dissolution data with missing elements.

The use of principal component analysis (PCA) for incomplete dissolution data sets is examined. The PC space is constructed using a reference set and the test set is projected in that space. Several cases such as a reference set with missing data, an incomplete test set and both sets measured at different time points, are discussed using two examples: one simulation and one obtained from the pharmaceutical practice. From the many possibilities to deal with missing data, the expectation-maximization algorithm in combination with PCA was chosen. The influence on the similarity or f2 factor is examined too. The sampling with replacement or bootstrap technique, which can be used to obtain confidence limits, can also be used when missing data are present in one of the data sets.

Algorithms↗
PubMed2002

Exploratory analysis of data sets with missing elements and outliers.

The main goal of the presented paper was to develop a general strategy allowing exploration of contaminated data sets with missing elements, based on application of robust PLS for initial estimation of missing elements. Using robust distance, the outlying elements were identified. After their identification and replacing by missing elements, the expectation-maximization algorithm (which can be built in into different computational procedures, such as principal component analysis and its generalisation to the N-way data-the TUCKER3 model) was used for construction of the final model.

Acid Rain↗
PubMed2001

The use of wavelets for signal denoising in capillary electrophoresis.

The discrete wavelet transform was applied to denoise electropherograms in capillary electrophoresis (CE). The use of the Haar wavelet and translation invariant denoising were found to be very efficient for this purpose. An important improvement was obtained, as compared with Savitzky-Golay and Fourier, which are the most commonly used techniques for denoising in the instrumentation software packages. A better removal of the noise and, especially, a better preservation of the shapes of very sharp peaks was achieved. Removal of the baseline variations was also investigated.

Journal Article↗
PubMed2000

The comparative molecular surface analysis (COMSA): a novel tool for molecular design.

A new method allowing for 3-D QSAR analysis and the prediction of biological activity is presented. Unlike comparative molecular field analysis (CoMFA)-like techniques, it is based not on a comparison of the properties characterizing a discrete set of points but on the mean electrostatic potential (MEP) calculated and labeling specific areas defined on the molecular surface. A Kohonen self-organizing neural network and partial least square (PLS) analysis have been used for performing such an operation. The series of steroids complexing the corticosteroid (CBG) and testosterone (TBG) globulins, which forms a benchmark measuring the performance of the methods in molecular design, and a series of benzoic acids described by the Hammett sigma constants is used for testing the method. It is demonstrated that a method can be used efficiently to evaluate the responses determined both by the combination of electrostatic and steric effects or by electrostatic effects alone, therefore, two different schemes were developed. The first one, which involves PLS analysis of the full comparative networks, covers both steric and electrostatic effects. This scheme works well for both the CBG and TBG data. The second scheme takes into account only the properties (MEP) of these regions within molecules that can be superimposed with the template molecule. This scheme provides the best predictive power for the benzoic acids series. Comparison of the results from a CoMFA analysis proves that method is at least as effective for the responses limited by electrostatic effects, although it significantly outperforms CoMFA for CBG affinity which is dominated by steric effects.

Benzoates↗
PubMed2000

Use of mass spectrometry for assessing similarity/diversity of natural products with unknown chemical structures.

An evaluation whether mass spectral data contain useful information for assessing similarity/diversity of drug compounds is presented. A comparative study was carried out between Ward's hierarchical agglomerative clustering, based on the 2D Daylight fingerprints or on the mass spectra, of a small database of 66 synthetic substances. The influence of normalization of the mass spectral data on the clustering result has also been studied. The results were subsequently compared with an expert's classification of the same small dataset, based on own evaluation according to known structure and pharmacological activity.

Cluster Analysis↗
PubMed2000

Self-organizing neural networks for modeling 3D QSAR--a comparative study.

Different architectures of self-organizing neural networks (SOM) have been used for modeling 3D QSAR. The atomic coordinates and partial atomic charges were used as input signals. In particular, the steroids complexing corticosteroid binding globulin (CBG) that are used as a benchmark measuring the performance of drug design methods have been applied to compare between individual methods. The sensitivity of the different architectures for changes of the alignments of the molecules within series, as well as the possibility for alignments based on the molecular inertial axes have been tested.

Drug Design↗
PubMedDate not supplied

Looking for natural patterns in analytical data. 2. Tracing local density with OPTICS.

The main principles and the algorithm of a density-based clustering approach, OPTICS, are described, and its unique properties, such as the ability to reveal clusters of arbitrary shapes and different densities, are illustrated on simulated and real spectral and chromatographic data sets. A "reachability plot" visualizing density fluctuations of data in multivariate space and a "color map" relating the original and/or descriptive features with data clustering allow a deeper insight into the data structure and its interpretation in chemical terms.

Journal Article↗
PubMedDate not supplied

Feature based fuzzy matching of 2D gel electrophoresis images.

Automatic alignment (matching) of two-dimensional gel electrophoresis images is of primary interest in the evolving field of proteomics. In the present study, feature-based matching techniques, in their classical and robust versions, are described, and an automatic method of fuzzy alignment (FA) is introduced. This method allows automatic matching of two gel images with different numbers of features with unknown correspondence. Performance of FA is tested on simulated and real data sets.

Automation↗
PubMedDate not supplied

On the optimal partitioning of data with K-means, growing K-means, neural gas, and growing neural gas.

In this paper, the performance of new clustering methods such as Neural Gas (NG) and Growing Neural Gas (GNG) is compared with the K-means method for real and simulated data sets. Moreover, a new algorithm called growing K-means, GK, is introduced as the alternative to Neural Gas and Growing Neural Gas. It has small input requirements and is conceptually very simple. The GK leads to nearly optimal values of the cost function, and, contrary to K-means, it is independent of the initial data set partition. The incremental property of GK additionally helps to estimate the number of "natural" clusters in data, i.e., the well-separated groups of objects in the data space.

Journal Article↗
PubMedDate not supplied

Matching 2D gel electrophoresis images.

Automatic alignment (matching) of two-dimensional gel electrophoresis images is of primary interest in the field of proteomics. The proposed method of 2D gel image matching is based on fuzzy alignment of features, extracted from gels' images, and it allows both global and local interpolation of image grid, followed by brightness interpolation. Method performance is tested on simulated images and gel images available via the Internet databases.

Algorithms↗