M. Daszykowski

Looking for natural patterns in data: Part 1. Density-based approach

Abstract A density-based unsupervised clustering approach for detecting natural patterns in data (further denoted as NP) is presented, and its performance is illustrated for data sets with different types of clusters. NP works for arbitrary clusters, is a single-scan technique, requires no presumptions regarding data distribution and requires only one input parameter, which describes the minimal number of objects, considered as cluster. Moreover, a comparison of NP with partitioning approaches is demonstrated. NP can be applied not only for data clustering, but also for the identification of outliers.

Representative subset selection

Fast development of analytical techniques enable to acquire huge amount of data. Large data sets are difficult to handle and therefore, there is a big interest in designing a subset of the original data set, which preserves the information of the original data set and facilitates the computations. There are many subset selection methods and their choice depends on the problem at hand. The two most popular groups of subset selection methods are uniform designs and cluster-based designs. Among the methods considered in this paper there are uniform designs, such as those proposed by Kennard and Stone, OptiSim, and cluster-based designs applying K-means technique and density based spatial clustering of applications with noise (DBSCAN). Additionally, a new concept of the subset selection with K-means is introduced.

Projection methods in chemistry

Abstract Visualization of a data set structure is one of the most challenging goals in data mining. Often, chemical data sets are multidimensional, and therefore visualization of their structure is not directly possible. To overcome this problem, the original data is compressed to the few new features by using projection techniques, preserving the original data structure as good as possible, and allowing its visualization. In this paper, a survey of different projection techniques, linear and nonlinear, is given. Their performance is illustrated on chemical data sets, and the advantages and disadvantages are pointed out.

Dealing with missing values and outliers in principal component analysis

An efficient methodology for dealing with missing values and outlying observations simultaneously in principal component analysis (PCA) is proposed. The concept described in the paper consists of using a robust technique to obtain robust principal components combined with the expectation maximization approach to process data with missing elements. It is shown that the proposed strategy works well for highly contaminated data containing different amounts of missing elements. The authors come to this conclusion on the basis of the results obtained from a simulation study and from analysis of a real environmental data set.

Determining orthogonal chromatographic systems prior to the development of methods to characterise impurities in drug substances.

To define starting conditions for the development of methods to separate impurities from the active substance and from each other in drugs with an unknown impurity profile, the parallel application of generic orthogonal chromatographic systems could be useful. The possibilities to define orthogonal chromatographic systems were examined by calculation of the correlation coefficients between retention factors k for a set of 68 drugs on 11 systems, by visual evaluation of the selectivity differences, by using principal component analysis, by drawing color maps and evaluating dendrograms. A zirconia-based stationary phase coated with a polybutadiene (PBD) polymer and three silica-based phases (base-deactivated, polar-embedded and monolithic) were used. Besides the stationary phase, the influence of pH and of organic modifier, on the selectivity of a system were evaluated. The dendrograms of hierarchical clusters were found good aids to assess orthogonality of chromatographic systems. The PBD-zirconia phase/methanol/pH 2.5 system is found most orthogonal towards several silica-based systems, e.g. a base-deactivated C16 -amide silica/methanol/pH 2.5 system. The orthogonality was validated using cross-validation, and two other validation sets, i.e. a set of non-ionizable solutes and a mixture of a drug and its impurities.

Metabolomics provide new insights on lung cancer staging and discrimination from chronic obstructive pulmonary disease.

Chronic obstructive pulmonary disease (COPD) and lung cancer are widespread lung diseases. Cigarette smoking is a high risk factor for both the diseases. COPD may increase the risk of developing lung cancer. Thus, it is crucial to be able to distinguish between these two pathological states, especially considering the early stages of lung cancer. Novel diagnostic and monitoring tools are required to properly determine lung cancer progression because this information directly impacts the type of the treatment prescribed. In this study, serum samples collected from 22 COPD and 77 lung cancer (TNM stages I, II, III, and IV) patients were analyzed. Then, a collection of NMR metabolic fingerprints was modeled using discriminant orthogonal partial least squares regression (OPLS-DA) and further interpreted by univariate statistics. The constructed discriminant models helped to successfully distinguish between the metabolic fingerprints of COPD and lung cancer patients (AUC training=0.972, AUC test=0.993), COPD and early lung cancer patients (AUC training=1.000, AUC test=1.000), and COPD and advanced lung cancer patients (AUC training=0.983, AUC test=1.000). Decreased acetate, citrate, and methanol levels together with the increased N-acetylated glycoproteins, leucine, lysine, mannose, choline, and lipid (CH3-(CH2)n-) levels were observed in all lung cancer patients compared with the COPD group. The evaluation of lung cancer progression was also successful using OPLS-DA (AUC training=0.811, AUC test=0.904). Based on the results, the following metabolite biomarkers may prove useful in distinguishing lung cancer states: isoleucine, acetoacetate, and creatine as well as the two NMR signals of N-acetylated glycoproteins and glycerol.

Classification and regression trees--studies of HIV reverse transcriptase inhibitors.

In this paper, the application of Classification And Regression Trees (CART) is presented for the analysis of biological activity of Non-Nucleoside Reverse Transcriptase Inhibitors (NNRTIs). The data consist of the biological activities, expressed as pIC50, of 208 NNRTIs against wild-type HIV virus (HIV-1) and four mutant strains (181C, 103N, 100I, 188L) and the computed interaction energies with the Reverse Transcriptase (RT) binding pocket. CART explains the observed biological activity of NNRTIs in terms of interactions with individual amino acids in the RT binding pocket, i.e., the original data variables.

Impurity fingerprints for the identification of counterfeit medicines - a feasibility study

Most of the counterfeit medicines are manufactured in non good manufacturing practices (GMP) conditions by uncontrolled or street laboratories. Their chemical composition and purity of raw materials may, therefore, change in the course of time. The public health problem of counterfeit drugs is mostly due to this qualitative and quantitative variability in their formulation and impurity profiles. In this study, impurity profiles were treated like fingerprints representing the quality of the samples. A total of 73 samples of counterfeit and imitations of Viagra(®) and 44 samples of counterfeit and imitations of Cialis(®) were analysed on a HPLC-UV system. A clear distinction has been obtained between genuine and illegal tablets by the mean of a discriminant partial least squares analysis of the log transformed chromatograms. Following exploratory analysis of the data, two classification algorithms were applied and compared. In our study, the k-nearest neighbour classifier offered the best performance in terms of correct classification rate obtained with cross-validation and during external validation. For Viagra(®), both cross-validation and external validation sets returned a 100% correct classification rate. For Cialis(®) 92.3% and 100% correct classification rates were obtained from cross-validation and external validation, respectively.

The Proteomic Analysis of Primary Cortical Astrocyte Cell Culture after Morphine Administration

Astrocytes are supportive cells, necessary for ensure optimal environment for neural cells functioning. They are involved in extracellular K+ level regulation and neurotransmitters removal. They are also dependent for myelination and synapses formation. They may make a contribution in signal propagation in the central nervous system, for example, through Ca2+ signaling. With the use of neonatal pure astrocyte cell culture, we investigated changes in astrocytes proteomes under the influence of morphine. We found 10 major proteins, which show different expression between physiological cell culture and morphine treatment. With 2D gel electrophoresis and nanoLC-ESI-MS/MS, we identified proteins and characterized their potential role in morphine dependence. Observed differences were also confirmed by Western blotting. Our data suggests a role for astrocytes in the formation of the morphine dependence at the molecular level. This finding may support interpretation of causes of morphine dependence formation based only on behavioral data.

A journey into low-dimensional spaces with autoassociative neural networks

The compression and the visualization of the data have been always a subject of a great deal of excitement. Since multidimensional data sets are difficult to interpret and visualize, much of the attention is drawn how to compress them efficiently. Usually, the compression of dimensionality is considered as the first step of exploratory data analysis. Here, we focus our attention on autoassociative neural networks (ANNs), which in a very elegant manner provide data compression and visualization. ANNs can deal with linear and nonlinear correlation among variables, what makes them a very powerful tool in exploratory data analysis. In the literature, ANNs are often referred as nonlinear principal component analysis (PCA), and due to their specific structure they are also known as bottleneck neural networks. In this paper, ANNs are discussed in details. Different training modes are described and illustrated on real example. The usefulness of ANNs for nonlinear data compression and visualization purposes is proven with the aid of chemical data sets, being the subject of analysis. The comparison of ANNs with well-known PCA is also presented.