Integrating genetic, functional genomic, and bioinformatics data in a systems biology approach to complex diseases: application to schizophrenia.

The search for DNA alterations that cause human disease has been an area of active research for more than 50 years, since the time that the genetic code was first solved. In the absence of data implicating chromosomal aberrations, researchers historically have performed whole genome linkage analysis or candidate gene association analysis to develop hypotheses about the genes that most likely cause a specific phenotype or disease. Whereas whole genome linkage analysis examines all chromosomal locations without a priori predictions regarding what genes underlie susceptibility, candidate gene association studies require a researcher to know in advance the genes that he or she wishes to test (based on their knowledge of a disease). To date, very few whole genome linkage studies and candidate gene studies have produced results that lead to generalizable findings about common diseases. One factor contributing to this lack of results has certainly been the previously limited resolution of the techniques. Recent technological advances, however, have made it possible to perform highly informative whole genome linkage and association analyses, as well as whole genome transcription (transcriptome) analysis. In addition, for the first time we can detect structural DNA aberrations throughout the genome on a fine scale. Each of these four approaches has its own strengths and weaknesses, but taken together, the results from an integrated analysis can implicate highly promising novel candidate genes. Here, we provide an overview of the integrated methodology that we have used to combine high-throughput genetic and functional genomic data with bioinformatics data that have produced new insights into the potential biological basis for schizophrenia. We believe that the potential of this combined approach is greater than that of a single mode of discovery, particularly for complex genetic diseases.

Quantitative analysis of dimethyl titanocene by iodometric titration, gas chromatography and NMR.

In this study we report the use of an automated iodometric titration method and a novel gas chromatography (GC) method for the quantitative analysis of dimethyl titanocene (DMT), a key raw material in drug synthesis. Both approaches are based on the reaction of DMT in toluene or tetrahydrofuran solutions with iodine. In the case of iodometric titration, excess iodine is titrated with a standardized aqueous sodium thiosulfate solution to a potentiometric end-point for the determination of DMT concentration. Alternatively, GC is employed to measure the concentration of iodomethane, a product of the reaction between DMT and iodine, in order to determine the concentration of DMT in the solution. Excellent agreement between iodometric titration, GC and NMR results using several DMT samples confirms the accuracy of the two methods and strongly supports the use of either method as a replacement to the expensive NMR for quantitative DMT analysis. The relatively few sources of error associated with the two methods, their ubiquitous nature and ease of application in routine analysis make them the analytical methods of choice, among all. Both methods have been validated according to ICH requirements. The use of iodometric titration method for DMT analysis is demonstrated with a couple of applications.

Image classification for content-based indexing.

Grouping images into (semantically) meaningful categories using low-level visual features is a challenging and important problem in content-based image retrieval. Using binary Bayesian classifiers, we attempt to capture high-level concepts from low-level image features under the constraint that the test image does belong to one of the classes. Specifically, we consider the hierarchical classification of vacation images; at the highest level, images are classified as indoor or outdoor; outdoor images are further classified as city or landscape; finally, a subset of landscape images is classified into sunset, forest, and mountain classes. We demonstrate that a small vector quantizer (whose optimal size is selected using a modified MDL criterion) can be used to model the class-conditional densities of the features, required by the Bayesian methodology. The classifiers have been designed and evaluated on a database of 6931 vacation photographs. Our system achieved a classification accuracy of 90.5% for indoor/outdoor, 95.3% for city/landscape, 96.6% for sunset/forest and mountain, and 96% for forest/mountain classification problems. We further develop a learning method to incrementally train the classifiers as additional data become available. We also show preliminary results for feature reduction using clustering techniques. Our goal is to combine multiple two-class classifiers into a single hierarchical classifier.

Enantiomeric separation of some pharmaceutical intermediates and reversal of elution orders by high-performance liquid chromatography using cellulose and amylose tris(3,5-dimethylphenylcarbamate) derivatives as stationary phases.

Several pairs of enantiomers of pharmaceutical intermediates were separated by HPLC directly on cellulose and amylose tris(3,5-dimethylphenylcarbamate) derivatives (Chiralcel OD and Chiralpak AD) using hexane as mobile phase with 2-propanol or ethanol as modifier. The separation and elution order of the enantiomers on the two columns using different alcohol modifiers were compared. Reversal of the elution order of some enantiomeric pairs associated with increased retention of many of these solutes upon changing the mobile phase modifier from 2-propanol to ethanol was observed. The effect of structural variation of two pairs of enantiomers on their k' and separation factor alpha was noted. Chiralcel OD and Chiralpak AD columns provided different retention, separation and elution order of some of the enantiomeric pairs.

Retention in reversed-phase chromatography: partition or adsorption?

A unified framework within the hermeneutics of the solvophobic theory is employed for the treatment of experimental data with nonpolar and weakly polar substances in reversed-phase chromatography (RPC), oil-water partitioning and adsorption on activated charcoal from dilute aqueous solution. This approach sheds light on the energetic similarities between such processes driven by the hydrophobic effect. Among several stationary phase models that have been proposed in the literature for the physical representation of alkyl-silica bonded phases, the isolated solvated hydrocarbon chains model is adopted for the retention in RPC since it represents most closely the stationary phase configuration and is not based a priori on a partition or adsorption mechanism as some other models are for the retention in RPC. Using the fundamental framework of the solvophobic theory, the free energy change per unit nonpolar surface area for octanol-water and hexadecane-water partitioning, retention in RPC as well as adsorption on activated charcoal from dilute aqueous solution at 25 degrees C are evaluated and they are found to be in good agreement with the corresponding experimental data. Furthermore, such quantities are very similar for all the above mentioned processes involving aqueous solution, in contradistinction to the predictions by the lattice theory. From the results it follows that these apparently disparate processes are subject to the same physicochemical principle. The present study demonstrates the capability of the solvophobic theory in describing the energetics of processes involving hydrophobic interactions, and exposes the difficulties in distinguishing between partition and adsorption mechanisms in RPC by using partition models based on the lattice approach. It is concluded that a clear distinction between partition and adsorption in RPC of nonpolar elites is not apparent from thermodynamic analysis.

Retention in hydrophobic interaction chromatography and dissolution of nonpolar gases in water.

The retention data of dansyl amino acids in hydrophobic interaction chromatography (HIC) is interpreted in terms of physicochemical properties of the eluites and solvent and the results are compared to data for the dissolution in water of nonpolar gases. Isothermodynamic temperatures, certain linear relationships between thermodynamic quantities and molecular structure as well as the hermeneutics of the solvophobic theory are employed for data analysis. van't Hoff plots of the retention data with dansyl amino acids are curved and intersect at the isoenergetic temperature, indicative of enthalpy-entropy compensation. The isoenthalpic and isoentropic temperatures are also evaluated and the three isothermodynamic temperatures are found to fall in the same narrow ranges as those for the dissolution of gases in water. Plots of thermodynamic quantities for the retention in HIC against the nonpolar molecular area are linear and thus allow the evaluation of enthalpy, entropy, Gibbs energy and heat capacity changes per unit nonpolar surface area of dansyl amino acids. By employing the solvophobic theory thermodynamic expressions are derived in terms of the nonpolar molecular area and interfacial tension. Using these expressions and n-heptane as model compound it was found that the above group molecular parameters are nearly identical for the retention in HIC and the dissolution of gases in water, thus confirming the mechanistic identity of the two processes.

Temperature effects in hydrophobic interaction chromatography.

The effect of temperature from 5 degrees C to 50 degrees C on the retention of dansyl derivatives of amino acids in hydrophobic interaction chromatography (HIC) was investigated by HPLC on three stationary phases. Plots of the logarithmic retention factor against the reciprocal temperature in a wide range were nonlinear, indicative of a large negative heat capacity change associated with retention. By using Kirchoff's relations, the enthalpy, entropy, and heat capacity changes were evaluated from the logarithmic retention factor at various temperatures by fitting the data to a logarithmic equation and a quadratic equation that are based on the invariance and on an inverse square dependence of the heat capacity on temperature, respectively. In the experimental temperature interval, the heat capacity change was found to increase with temperature and could be approximated by the arithmetic average. For HIC retention of a set of dansylamino acids, both enthalpy and entropy changes were positive at low temperatures but negative at high temperatures as described in the literature for other processes based on the hydrophobic effect. The approach presented here shows that chromatographic measurements can be not only a useful adjunct to calorimetry but also an alternative means for the evaluation of thermodynamic parameters.