Dayanand N. Naik

Computational protein biomarker prediction: a case study for prostate cancer

BackgroundRecent technological advances in mass spectrometry pose challenges in computational mathematics and statistics to process the mass spectral data into predictive models with clinical and biological significance. We discuss several classification-based approaches to finding protein biomarker candidates using protein profiles obtained via mass spectrometry, and we assess their statistical significance. Our overall goal is to implicate peaks that have a high likelihood of being biologically linked to a given disease state, and thus to narrow the search for biomarker candidates.ResultsThorough cross-validation studies and randomization tests are performed on a prostate cancer dataset with over 300 patients, obtained at the Eastern Virginia Medical School using SELDI-TOF mass spectrometry. We obtain average classification accuracies of 87% on a four-group classification problem using a two-stage linear SVM-based procedure and just 13 peaks, with other methods performing comparably.ConclusionsModern feature selection and classification methods are powerful techniques for both the identification of biomarker candidates and the related problem of building predictive models from protein mass spectrometric profiles. Cross-validation and randomization are essential tools that must be performed carefully in order not to bias the results unfairly. However, only a biological validation and identification of the underlying proteins will ultimately confirm the actual value and power of any computational predictions.

Analysis of multivariate repeated measures data with a Kronecker product structured covariance matrix

In this article we consider a set of t repeated measurements on p variables (or characteristics) on each of the n individuals. Thus, data on each individual is a p 2 t matrix. The n individuals themselves may be divided and randomly assigned to g groups. Analysis of these data using a MANOVA model, assuming that the data on an individual has a covariance matrix which is a Kronecker product of two positive definite matrices, is considered. The well-known Satterthwaite type approximation to the distribution of a quadratic form in normal variables is extended to the distribution of a multivariate quadratic form in multivariate normal variables. The multivariate tests using this approximation are developed for testing the usual hypotheses. Results are illustrated on a data set. A method for analysing unbalanced data is also discussed.

Creation of environmental health information system for public health service: A pilot study

Despite more than a decade of research on medical information systems, deficiencies exist in our capability of establishing an effective environmental health information infrastructure. In this research, we present a pilot study on creating a feasible environmental health information infrastructure. The newly-developed environmental health information system is a web-based platform that integrates databases, decision-making tools, geographic information systems for supporting public health service and policy making. The study, which is a part of a comprehensive effort known as Environmental Public Health Tracking proposed by the Center for Disease Control and Prevention, opens the door for future research on a large scale nation-wide healthcare information infrastructure.

Analysis of multivariate longitudinal data using quasi-least squares

In this paper we consider the analysis of multivariate longitudinal data assuming a scale multiple of Kronecker product correlation structure for the covariance matrix of the observations on each subject. The method used for the estimation of the parameters is the quasi-least squares method developed in the following three papers: Chaganty (J. Statist. Plann. Inference 63 (1997) 39), Shults and Chaganty (Biometrics 54 (1998) 1622) and Chaganty and Shults (J. Statist. Plann. Inference 76 (1999) 145). We show that the estimating equations for the correlation parameters in the quasi-least-squares method are optimal unbiased estimating equations if the data is from a normal population. An algorithm for computing the estimates is provided and implemented on a real life data set. The asymptotic joint distribution of the estimators of the regression and correlation parameters is derived and used for testing a linear hypothesis on the regression parameters.

Detection of influential observations in multivariate regression

Several methods have been suggested, in the literature, to detect influential observations from the data fitting usual linear model y=X∗∗∗+∗∗∗, ∗∗∗∽N(0, ∗∗∗2I). Recently, Chatterjee & Hadi (1986) have reviewed most of these available methods and described the inter-relationships between them. In this article, we extend some of these methods to the case of multivariate regression data. We consider several data sets to illustrate the methods.

Revisiting Olympic Track Records: Some Practical Considerations in the Principal Component Analysis

Abstract In some practical problems where a principal component analysis is utilized, the use of the variance covariance matrix of an appropriately defined set of variables, rather than the correlation matrix, may be more meaningful. This is illustrated through the analysis of 1984 Olympic records data on various track events. The analysis results in conclusions that are more appealing to intuition and that are also consistent with a retrospective visual examination of the data on certain leading countries in their athletic excellence.

Andrews plots for multivariate data: some new suggestions and applications

Andrews plots (Biometrics 28 (1972) 125-136), as a tool to graphically interpret multivariate data, have recently gained considerable recognition. In this article, we first review the previous literature and then suggest a modification to the traditional Andrews plots. Finally, we illustrate a few new applications of these plots in robust design studies and in correspondence analysis, using real data.

Microwave-assisted synthesis, molecular docking and antitubercular activity of 1,2,3,4-tetrahydropyrimidine-5-carbonitrile derivatives.

Based on bioisosteric similarities with isoniazid, a series of 1,2,3,4-tetrahydropyrimidine-5-carbonitrile derivatives has been designed. The target compounds have been synthesized by multicomponent reaction which involves one-pot organic reactions using ethylcyanoacetate, urea/thiourea and arylaldehydes in presence of ethanolic K(2)CO(3). Two methodologies, conventional and microwave-assisted, have been adopted for the synthesis. The later strategy gave high yields in less than 10 min as compared to long hours using the former approach. Molecular docking of the target compounds into the enzyme Mycobacterium tuberculosis enoyl reductase (InhA) revealed important structural information on the plausible binding interactions. Major binding interactions were found to be of dispersion type (residues Tyr158, Ile215, Met103 and Met199) and a hydrogen bond with Tyr158. Binding poses of the all the compounds were energetically favorable and showed good interactions with the active site residues. Few selected compounds were also evaluated for antitubercular activity in vitro against drug-sensitive M. tuberculosis H37Rv strain and clinically isolated S, H, R and E resistant M. tuberculosis by luciferase reporter phage (LRP) assay method. Some compounds displayed promising antimycobacterial activity comparable or less than the standard drugs isoniazid and rifampicin.

Estimation of variance components in staggered nested designs

Variance components are estimated by two different methods for a general p stage random-effects staggered nested design. In addition to estimation from an analysis of variance, a new approach is introduced. The main features of this new technique are its simplicity and its ability to yield non-negative estimates of the variance components. The performances of the two procedures are compared using simulation and the meansquared-error criterion.

Influence of kinematic analysis methods on detecting ankle and subtalar joint instability.

Patients with subtalar joint instability may be misdiagnosed with ankle instability, which may lead to chronic instability at the subtalar joint. Therefore, it is important to understand the difference in kinematics after ligament sectioning and differentiate the changes in kinematics between ankle and subtalar instability. Three methods may be used to determine the joint kinematics; the Euler angles, the Joint Coordinate System (JCS) and the helical axis (HA). The purpose of this study was to investigate the influence of using either method to detect subtalar and ankle joints instability. 3D kinematics at the ankle and subtalar joint were analyzed on 8 cadaveric specimens while the foot was intact and after sequentially sectioning the anterior talofibular ligament (ATFL), the calcaneofibular ligament (CFL), the cervical ligament and the interosseous talocalcaneal ligament (ITCL). Comparison in kinematics calculated from sensor and anatomical landmarks was conducted as well as the influence of Euler angles and JCS rotation sequence (between ISB recommendation and previous research) on the subtalar joint. All data showed a significant increase in inversion when the ITCL was sectioned. There were differences in the data calculated using sensors coordinate systems vs. anatomic coordinate systems. Anatomic coordinate systems were recommended for these calculations. The Euler angle and JCS gave similar results. Differences in Euler angles and JCS sequence lead to the same conclusion in detecting instability at the ankle and subtalar joint. As expected, the HA detected instability in plantarflexion at the ankle joint and in inversion at the subtalar joint.