A new integrated statistical approach to the diagnostic use of two-dimensional maps

Two-dimensional (2-D) electrophoresis is a very useful technique for the analysis of proteins in biological tissues. The complexity of the 2-D maps obtained causes many difficulties in the comparison of different samples. A new method is proposed for comparing different 2-D maps, based on five steps: (i) the digitalisation of the image; (ii) the transformation of the digitalised map in a fuzzy entity, in order to consider the variability of the 2-D electrophoretic separation; (iii) the calculation of a similarity index for each pair of maps; (iv) the analysis by multidimensional scaling of the previously obtained similarity matrix; (v) the analysis by classification or cluster analysis techniques of the resulting map co-ordinates. The method adopted was first tested on some simulated samples in order to evaluate its sensitivity to small changes in the spots position and size. The optimal setting of the method parameters was also investigated. Finally, the method was successfully applied to a series of real samples corresponding to the electrophoretic bidimensional analysis of sera from normal and nicotine-treated rats. Multidimensional scaling allowed the separation of the two classes of samples without any misclassification.     

Novel characterization of proteomics maps by sequential neighborhoods of protein spots

We consider a characterization of proteomics maps based on an alternative kind of neighborhood graphs for the protein spots on 2-D gel. The novel approach considers for every protein spot only the nearest neighborhood consisting of protein spots of higher abundance. The approach has the simplicity and advantages of the recently introduced characterization of proteome maps based on considering the nearest neighborhoods of protein spots, but it also has important additional desirable computational features. The characterization of the nearest neighborhood graphs of 2-D gel proteomics maps is sensitive to the number of spots considered and may lead to changes in the degree of similarity of different maps when the number of points has been changed, thus imposing restrictions on the protocol used for comparison of maps. The novel approach presented in this work is less sensitive to the number of points used in the analysis because graphs are constructed in a stepwise process in which the role of more distant neighbors has been diminished by linking a new spot to the nearest spot that has been already part of the neighborhood graph. In this way a graph with N + 1 spots is obtained from the graph on N spots by adding a single new link, while in the case of the nearest neighborhood graphs adding a new spot introduces novel neighborhoods and generally results in a graph that may differ significantly from the neighborhood graph on N points.     

Comparison of sequences as a method for evaluation of the molecular similarity

String comparison techniques were developed and applied for measuring the molecular similarity of chemical structures. The molecular structures were encoded as a sequence of numbers representing counts of paths of different lengths. The similarity index between two compounds was calculated as the difference between the gains of information derived through comparison of the corresponding molecular path sequences. Ranks between the structures of the studied data base obtained according to this similarity were used as basic data for deriving correspondences between the elements of the set of compounds. The method was applied on a group of 41 barbiturates. Correlation equations were calculated for different groups of compounds grouped according to the displayed similarity. The correlation equations and the corresponding statistics were obtained using standard computer programs. Special algorithm for computing the similarity index and the correlation matrix (outlined very briefly) was developed and implemented on VAX 11/750.     

Proteomic analysis of cellular soluble proteins from human bronchial smooth muscle cells by combining nondenaturing micro 2DE and quantitative LC‐MS/MS. 2. Similarity search between protein maps for the analysis of protein complexes

Human bronchial smooth muscle cell soluble proteins were analyzed by a combined method of nondenaturing micro 2DE, grid gel‐cutting, and quantitative LC‐MS/MS and a native protein map was prepared for each of the identified 4323 proteins [1]. A method to evaluate the degree of similarity between the protein maps was developed since we expected the proteins comprising a protein complex would be separated together under nondenaturing conditions. The following procedure was employed using Excel macros; (i) maps that have three or more squares with protein quantity data were selected (2328 maps), (ii) within each map, the quantity values of the squares were normalized setting the highest value to be 1.0, (iii) in comparing a map with another map, the smaller normalized quantity in two corresponding squares was taken and summed throughout the map to give an “overlap score,” (iv) each map was compared against all the 2328 maps and the largest overlap score, obtained when a map was compared with itself, was set to be 1.0 thus providing 2328 “overlap factors,” (v) step (iv) was repeated for all maps providing 2328 × 2328 matrix of overlap factors. From the matrix, protein pairs that showed overlap factors above 0.65 from both protein sides were selected (431 protein pairs). Each protein pair was searched in a database (UniProtKB) on complex formation and 301 protein pairs, which comprise 35 protein complexes, were found to be documented. These results demonstrated that native protein maps and their similarity search would enable simultaneous analysis of multiple protein complexes in cells.     

Global mapping of rat plasma proteins with a native proteomic approach using nondenaturing micro 2DE and quantitative LC‐MS/MS

Plasma samples from adult male rats were separated by nondenaturing micro 2DE and a reference gel was selected, on which 136 CBB‐stained spots were numbered and subjected to in‐gel digestion and quantitative LC‐MS/MS. The analysis provided the assignment of 1–25 (average eight) non‐redundant proteins in each spot and totally 199 proteins were assigned in the 136 spots. About 40% of the proteins were detected in more than one spot and 15% in more than ten spots. We speculate this complexity arose from multiple causes, including protein heterogeneity, overlapping of protein locations and formation of protein complexes. Consequently, such results could not be appropriately presented as a conventional 2DE map, i.e. a list or a gel pattern with one or a few proteins annotated to each spot. Therefore, the LC‐MS/MS quantity data was used to reconstruct the gel distribution of each protein and a library containing 199 native protein maps was established for rat plasma. Since proteins that formed a complex would migrate together during the nondenaturing 2DE and thus show similar gel distributions, correlation analysis was attempted for similarity comparison between the maps. The protein pairs showing high correlation coefficients included some well‐known complexes, suggesting the promising application of native protein mapping for interaction analysis. With the importance of rat as the most commonly used laboratory animal in biomedical research, we expect this work would facilitate relevant studies by providing not only a reference library of rat plasma protein maps but a means for functional and interaction analysis.     

Profiling stage-dependent changes of protein expression in Caenorhabditis elegans by mass spectrometric proteome analysis leads to the identification of stage-specific marker proteins

Proteome maps obtained by synchronization of the wild-type Caenorhabditis elegans development reflected stage-dependent molecular differences and revealed dynamic cytoskeletal processes during ontogenesis. Distinct protein spots that may function as molecular markers for the corresponding developmental stages were mass spectrometrically identified. The amount of the Cu(2+)- Zn(2+) superoxide dismutase (CE23550) and an aspartyl proteinase (CE21681) was highest in the first larval stage (L1) and decreased during the ontogenesis from the first larval stage to the adult. Tropomyosin III (CE29059) was prominently present in the first and second larval stage (L1/L2). Abundances of actin 1 or 4 (CE12358 or CE13148) and tropomyosin I (CE28782) were particularly high in multiple spots in the third larval stage (L3). Interestingly, the amount of DIM-1 protein (CE27706), reflected by two spots, was the lowest in this stage. A particular splicing factor (CE31089) was detected only in the fourth larval stage (L4), whereas a spot with high abundance representing the cuticle collagen (CE02272) was only found highly expressed in adult animals (A). In addition, a Ca(2+)-binding protein (CE12368) and one protein spot which has not yet been identified, both reached their maximal spot intensities in the adult stage (A). Moreover, the ASP-1, CCT-5, GPD-1, GPD-2, HSP-6, HSP-16.2, IFB-2, LEC-2, LIN-53, LMN-1, MDH-1, NUD-1, RPA-0, RSP-12, SOD-1, TBB-1, TBB-2, TMY-1, UNC-60, and VIT-2 proteins for which mutants are available and two still unidentified protein spots which were present in all developmental stages, have been reproducibly localized in proteome maps of distinct ontogenesis states.     

Extending molecular similarity to energy surfaces: Boltzmann similarity measures and indices

Maxwell-Boltzmann statistics provides the adequate mathematical background allowing to define similarity measures involving molecular energy surfaces and electrostatic potential maps. Boltzmann similarity measures are described and various illustrative examples are used to show the practical viability of the theory. A new molecular similarity index is also presented. Finally, hybrid measures involving Boltzmann and density distributions are defined.     

Study of the similarity between distribution maps of concentration in near-infrared spectroscopy chemical imaging obtained by different multivariate calibration approaches

Nowadays, near-infrared spectroscopy chemical imaging (NIR-CI) has been widely used in pharmaceutical analysis since it provides important surface information about the samples. In this work the information of NIR-CI at the pixel level was compared through calculation of the similarity between distribution maps of concentration obtained by different multivariate calibration approaches. The comparison was performed by using four different multivariate methods (MCR, MLR, CLS and PLS) in analysis of carbamazepine pharmaceutical formulations. For global determination, all models developed showed RMSEP below 1.9% (w/w) for active principal ingredient (API) and better than 4.6% (w/w) for excipients. Also, the distribution maps obtained by PLS, CLS and MCR showed great similarity for all compounds of the formulation as well with concentrations in the tablets. However, comparing the distribution maps obtained by MLR with those from the other chemometric tools, a lower similarity was observed. Thus, this fitted model does not ensure, by itself, that the images obtained are reliable or accurate. The paper also compares the distribution maps of concentrations obtained from all constituents present in the pharmaceutical formulation with their respective micrographs.     

含呋喃环双酰脲类衍生物的三维定量构效关系研究

采用比较分子力场分析法(CoMFA)和比较分子相似性指数分析法(CoMSIA), 对27个新型双酰基脲类化合物的杀蚊幼虫(Aedes aegypti L.)活性进行三维定量构效关系(3D-QSAR)研究. 在CoMFA研究中, 考察了网格点步长对统计结果的影响. 在CoMSIA研究中, 系统考察了各种分子场组合、网格点步长和衰减因子对模型统计结果的影响, 发现立体场和氢键供体场的组合得到最佳模型. 所建立的CoMFA和CoMSIA模型的非交叉验证相关系数r2值分别为0.828和0.841, 并都具有较强的预测能力. CoMFA和CoMSIA模型的三维等值图不仅直观地解释了结构与活性的关系, 而且为后续优化该系列化合物提供了理论依据.     

Spot overlapping in two-dimensional polyacrylamide gel electrophoresis separations: a statistical study of complex protein maps

A statistical approach able to extract the information contained in a two-dimenisional polyacrylamide gel electrophoresis (2-D PAGE) separation is here reported. The method is based on the quantitative theory of peak overlapping, a procedure previously developed by the authors and here extended to 2-D separations. The whole map is divided into many strips in order to obtain 1-D separations on which the statistic procedure is applied: the developed algorithms, on the basis of spot experimental data (intensity and spatial coordinates) permit to estimate the intrinsic number of components and to single out the specific order present in spot positions. The procedure was validated on computer-simulated maps. Its applicability to real samples was tested on maps obtained from literature sources. The following important information on protein mixtures can be extracted: (i) the number of proteins can be accurately estimated, on the basis of the spatial coordinates and intensities of spots detected in the 2-D PAGE map; (ii) the model describing distribution of interdistance between adjacent spots can be identified in both the separation dimensions; (iii) the presence of repeated interdistances in spot positions in the maps can be easily singled out: these regularities suggest specific protein modifications.     

Cumulative and discrete similarity analysis of electrostatic potentials and fields

A new index suitable for computing molecular similarity based upon the similarity of molecular properties such as electrostatic potentials or electrostatic fields is presented in two forms. For one form of the present index, general conditions are established for which a linear measure of similarity is obtained. An illustrative example is provided in which the electrostatic field and electrostatic potential of guanine obtained from different wave functions are compared. © 1993 John Wiley & Sons, Inc.     

Proteomic profiling of pancreatic ductal carcinoma cell lines treated with trichostatin-A

A pancreatic adenocarcinoma cell line (Paca44) was treated with trichostatin-A (TSA), a potent inhibitor of histone deacetylases, in order to evaluate the effect of this drug on protein expression. Master maps of control and treated Paca44 cells were generated by analysis with the PDQuest software. The comparison between such maps showed up- and downregulation of 51 polypeptide chains, out of a total of 700 spots detected by a medium-sensitivity stain, micellar Coomassie Brilliant Blue. Fingerprinting by matrix-assisted laser desorption/ionization-time of flight (MALDI-TOF)-mass spectrometry analysis enabled the identification of 22 of these spots. Among these proteins, of particular interest are the two downregulated proteins nucleophosmin and translationally controlled tumor protein, as well as the upregulated proteins programmed cell death protein 5 (also designated as TFAR19) and stathmin (oncoprotein 18). The modulation of these four proteins is consistent with our observation that TSA is able to inhibit cell growth of Paca44 by causing cell cycle arrest at the G2 phase and apoptotic cell death.     

Study of spectral analytical data using fingerprints and scaled similarity measurements

A new chemoinformatic model has been developed for enlarging the differences between spectra and applied to differentiation of wines according to the criteria grape origin and variety and ageing process. The model is based on generation of fingerprints from normalised spectra, using empirical parameters and a set of 120 samples. After generation of the fingerprints, similarity matrixes were built on the basis of the Tanimoto similarity index between the fingerprints of the samples. Calculation of the Tanimoto index was modified to adapt the index to the characteristics of the analytical measurements. Thus, scaling factors taking into account pattern fingerprints generated from a group of samples with common characteristics were used. In addition, a modified expression for calculating the Tanimoto index was employed. Principal-components analysis (PCA) and soft independent modelling of class analogy (SIMCA) were applied to the similarity matrixes. The results obtained are discussed as a function of the normalisation method employed, the empirical factor used in generation of the fingerprints, and selection of samples for building the pattern fingerprint, etc. Finally, results from differentiation of wines are compared with those obtained by applying PCA to the unprocessed spectra as stated by the proposed model.     

On 3-D graphical representation of proteomics maps and their numerical characterization

We consider numerical characterization of proteomics maps by representing a map as a three-dimensional graphical object based on x, y coordinates of the spots and using their relative abundance as the z coordinate. In our representation the protein spots are first ordered based on their relative abundance and labeled accordingly. In the next step a 3-D path is constructed connecting spots having adjacent labels. Finally a matrix is constructed by assigning to each pairs of labels (i, j) matrix element, the numerical value of which is based on the quotients of the Euclidean distance and the distance along the 3-D zigzag between the two points. The approach has been illustrated on a fragment of a proteomics map and compared with 2-D graphical representation of proteomics maps.     

Optimization of Carbó molecular similarity index using gradient methods

A steepest descent method for optimizing the Carbó molecular similarity index was implemented and evaluated. Comparisons were made between this procedure and the extensively used simplex method. Several data sets were considered, and in each case the gradient method showed a substantial improvement in the time taken for the optimization to converge while comparable similarity values were obtained. In some cases, performance enhancements of up to an order of magnitude were observed. © 1997 by John Wiley & Sons, Inc.     

Proteome analysis of human stomach tissue: separation of soluble proteins by two-dimensional polyacrylamide gel electrophoresis and identification by mass spectrometry

Two-dimensional gel electrophoresis (2-DE) maps for human stomach tissue proteins have been prepared by displaying the protein components of the tissue by 2-DE and identifying them using mass spectrometry. This will enable us to present an overview of the proteins expressed in human stomach tissues and lays the basis for subsequent comparative proteome analysis studies with gastric diseases such as gastric cancer. In this study, 2-DE maps of soluble fraction proteins were prepared on two gel images with partially overlapping pH ranges of 4-7 and 6-9. On the gels covering pH 4-7 and pH 6-9, about 900 and 600 protein spots were detected by silver staining, respectively. For protein identification, proteins spots on micropreparative gels stained with colloidal Coomassie Brilliant Blue G-250 were excised, digested in-gel with trypsin, and analyzed by peptide mass fingerprinting with delayed extraction-matrix assisted laser desorption/ionization-mass spectrometry (DE-MALDI-MS). In all, 243 protein spots (168 spots in acidic map and 75 spots in basic map) corresponding to 136 different proteins were identified. Besides these principal maps, overview maps (displayed on pH 3-10 gels) for total homogenate and soluble fraction, are also presented with some identifications mapped on them. Based on the 2-DE maps presented in this study, a 2-DE database for human stomach tissue proteome has been constructed and is available at http://proteome.gsnu.ac.kr/DB/2DPAGE/Stomach/. The 2-DE maps and the database resulting from this study will serve important resources for subsequent proteomic studies for analyzing the normal protein variability in healthy tissues and specific protein variations in diseased tissues.     

On characterizing proteomics maps by using weighted Voronoi maps

In contrast to the standard construction of Voronoi regions, in which the boundaries between different regions are at equal distance from the given points, we consider the construction of modified Voronoi regions obtained by giving greater weights to spots reported to have higher abundance. Specifically we are interested in applying this approach to 2-D proteomics maps and their numerical characterization. As will be seen, the boundaries of the weighted Voronoi regions are sensitive to the relative abundances of the protein spots and thus the abundances of protein spots, the z component of the (x, y, z) triplet, are automatically incorporated in the numerical analysis of the adjacency matrix, rather than used to augment the adjacency matrix as non-zero diagonal matrix elements. The outlined approach is general and it may be of interest for numerical analyses of other maps that are defined by triplets (x, y, z) as input information.     

A proteomic analysis of secreted proteins from xylan-induced Bacillus sp. strain K-1

The expression level of extracellular proteins in an alkaliphilic bacterium, Bacillus sp. strain K-1, grown in a xylan-containing medium, is significantly increased when compared with that grown in the nonxylan culture medium. A proteomic approach has been efficiently applied to separate and characterize these differentially expressed secretory proteins. Eight prominent protein spots were identified and subjected to N-terminal amino acid sequencing. The results show that three spots share considerable similarity with the xylanolytic enzymes and that two spots share considerable similarity with the GltC regulatory protein and 3-dehydroquinate dehydratase, respectively. In addition, the three other proteins show little similarity with the known proteins in the database. In conclusion, our results demonstrate that the proteomic approach is a highly efficient method to rapidly study the differential expression of the secreted proteins by Bacillus sp. strain K-1 grown under xylan-induced condition.