A Data-Analytic Strategy for Protein-Biomarker Discovery: Profiling of High-Dimensional Proteomic Data for Cancer Detection

With recent advances in mass spectrometry techniques, it is now possible to investigate proteins over a wide range of molecular weights in small biological specimens. This advance has generated data-analytic challenges in proteomics, similar to those created by microarray technologies in genetics, namely, discovery of "signature" protein profiles specific to each pathologic state (e.g., normal vs. cancer) or differential profiles between experimental conditions (e.g., treated by a drug of interest vs. untreated) from high-dimensional data. We propose a data analytic strategy for discovering protein biomarkers based on such high-dimensional mass-spectrometry data. A real biomarker-discovery project on prostate cancer is taken as a concrete example throughout the paper: the project aims to identify proteins in serum that distinguish cancer, benign hyperplasia, and normal states of prostate using the Surface Enhanced Laser Desorption/Ionization (SELDI) technology, a recently developed mass spectrometry technique. Our data analytic strategy takes properties of the SELDI mass-spectrometer into account: the SELDI output of a specimen contains about 48,000 (x, y) points where x is the protein mass divided by the number of charges introduced by ionization and y is the protein intensity of the corresponding mass per charge value, x, in that specimen. Given high coefficients of variation and other characteristics of protein intensity measures (y values), we reduce the measures of protein intensities to a set of binary variables that indicate peaks in the y-axis direction in the nearest neighborhoods of each mass per charge point in the x-axis direction. We then account for a shifting (measurement error) problem of the x-axis in SELDI output. After these pre-analysis processing of data, we combine the binary predictors to generate classification rules for cancer, benign hyperplasia, and normal states of prostate. Our approach is to apply the boosting algorithm to select binary predictors and construct a summary classifier. We empirically evaluate sensitivity and specificity of the resulting summary classifiers with a test dataset that is independent from the training dataset used to construct the summary classifiers. The proposed method performed nearly perfectly in distinguishing cancer and benign hyperplasia from normal. In the classification of cancer vs. benign hyperplasia, however, an appreciable proportion of the benign specimens were classified incorrectly as cancer. We discuss practical issues associated with our proposed approach to the analysis of SELDI output and its application in cancer biomarker discovery.

Percutaneous coronary interventions in Europe: prevalence, numerical estimates, and projections based on data up to 2004

AIMS: A registry mandated by the European Society of Cardiology collects data on trends in interventional cardiology within Europe. Special interest focuses on relative increases and ratios in new techniques and their distributions across Europe. We report the data through 2004 and give an overview of the development of coronary interventions since the first data collection in 1992. METHODS AND RESULTS: Questionnaires were distributed yearly to delegates of all national societies of cardiology represented in the European Society of Cardiology. The goal was to collect the case numbers of all local institutions and operators. The overall numbers of coronary angiographies increased from 1992 to 2004 from 684 000 to 2 238 000 (from 1250 to 3930 per million inhabitants). The respective numbers for percutaneous coronary interventions (PCIs) and coronary stenting procedures increased from 184 000 to 885 000 (from 335 to 1550) and from 3000 to 770 000 (from 5 to 1350), respectively. Germany was the most active country with 712 000 angiographies (8600), 249 000 angioplasties (3000), and 200 000 stenting procedures (2400) in 2004. The indication has shifted towards acute coronary syndromes, as demonstrated by rising rates of interventions for acute myocardial infarction over the last decade. The procedures are more readily performed and perceived safer, as shown by increasing rate of "ad hoc" PCIs and decreasing need for emergency coronary artery bypass grafting (CABG). In 2004, the use of drug-eluting stents continued to rise. However, an enormous variability is reported with the highest rate in Switzerland (70%). If the rate of progression remains constant until 2010 the projected number of coronary angiographies will be over three million, and the number of PCIs about 1.5 million with a stenting rate of almost 100%. CONCLUSION: Interventional cardiology in Europe is ever expanding. New coronary revascularization procedures, alternative or complementary to balloon angioplasty, have come and gone. Only stenting has stood the test of time and matured to the default technique. Facilitated access to PCI, more complete and earlier detection of coronary artery disease promise continued growth of the procedure despite the uncontested success of prevention.

A graph theoretic approach to testing associations between disparate sources of functional genomic data

The last few years have seen the advent of high-throughput technologies to analyze various properties of the transcriptome and proteome of several organisms. The congruency of these different data sources, or lack thereof, can shed light on the mechanisms that govern cellular function. A central challenge for bioinformatics research is to develop a unified framework for combining the multiple sources of functional genomics information and testing associations between them, thus obtaining a robust and integrated view of the underlying biology. We present a graph theoretic approach to test the significance of the association between multiple disparate sources of functional genomics data by proposing two statistical tests, namely edge permutation and node label permutation tests. We demonstrate the use of the proposed tests by finding significant association between a Gene Ontology-derived "predictome" and data obtained from mRNA expression and phenotypic experiments for Saccharomyces cerevisiae. Moreover, we employ the graph theoretic framework to recast a surprising discrepancy presented in Giaever et al. (2002) between gene expression and knockout phenotype, using expression data from a different set of experiments.

Bayesian Hidden Markov Modeling of Array CGH Data

Genomic alterations have been linked to the development and progression of cancer. The technique of Comparative Genomic Hybridization (CGH) yields data consisting of fluorescence intensity ratios of test and reference DNA samples. The intensity ratios provide information about the number of copies in DNA. Practical issues such as the contamination of tumor cells in tissue specimens and normalization errors necessitate the use of statistics for learning about the genomic alterations from array-CGH data. As increasing amounts of array CGH data become available, there is a growing need for automated algorithms for characterizing genomic profiles. Specifically, there is a need for algorithms that can identify gains and losses in the number of copies based on statistical considerations, rather than merely detect trends in the data. We adopt a Bayesian approach, relying on the hidden Markov model to account for the inherent dependence in the intensity ratios. Posterior inferences are made about gains and losses in copy number. Localized amplifications (associated with oncogene mutations) and deletions (associated with mutations of tumor suppressors) are identified using posterior probabilities. Global trends such as extended regions of altered copy number are detected. Since the posterior distribution is analytically intractable, we implement a Metropolis-within-Gibbs algorithm for efficient simulation-based inference. Publicly available data on pancreatic adenocarcinoma, glioblastoma multiforme and breast cancer are analyzed, and comparisons are made with some widely-used algorithms to illustrate the reliability and success of the technique.

On the Accelerated Failure Time Model for Current Status and Interval Censored Data

This paper introduces a novel approach to making inference about the regression parameters in the accelerated failure time (AFT) model for current status and interval censored data. The estimator is constructed by inverting a Wald type test for testing a null proportional hazards model. A numerically efficient Markov chain Monte Carlo (MCMC) based resampling method is proposed to simultaneously obtain the point estimator and a consistent estimator of its variance-covariance matrix. We illustrate our approach with interval censored data sets from two clinical studies. Extensive numerical studies are conducted to evaluate the finite sample performance of the new estimators.

A Diagnostic Test for the Mixing Distribution in a Generalised Linear Mixed Model

We introduce a diagnostic test for the mixing distribution in a generalised linear mixed model. The test is based on the difference between the marginal maximum likelihood and conditional maximum likelihood estimates of a subset of the fixed effects in the model. We derive the asymptotic variance of this difference, and propose a test statistic that has a limiting chi-square distribution under the null hypothesis that the mixing distribution is correctly specified. For the important special case of the logistic regression model with random intercepts, we evaluate via simulation the power of the test in finite samples under several alternative distributional forms for the mixing distribution. We illustrate the method by applying it to data from a clinical trial investigating the effects of hormonal contraceptives in women.

RANDOM EFFECTS MODELS IN A META-ANALYSIS OF THE ACCURACY OF DIAGNOSTIC TESTS WITHIN A GOLD STANDARD IN THE PRESENCE OF MISSING DATA

In evaluating the accuracy of diagnosis tests, it is common to apply two imperfect tests jointly or sequentially to a study population. In a recent meta-analysis of the accuracy of microsatellite instability testing (MSI) and traditional mutation analysis (MUT) in predicting germline mutations of the mismatch repair (MMR) genes, a Bayesian approach (Chen, Watson, and Parmigiani 2005) was proposed to handle missing data resulting from partial testing and the lack of a gold standard. In this paper, we demonstrate an improved estimation of the sensitivities and specificities of MSI and MUT by using a nonlinear mixed model and a Bayesian hierarchical model, both of which account for the heterogeneity across studies through study-specific random effects. The methods can be used to estimate the accuracy of two imperfect diagnostic tests in other meta-analyses when the prevalence of disease, the sensitivities and/or the specificities of diagnostic tests are heterogeneous among studies. Furthermore, simulation studies have demonstrated the importance of carefully selecting appropriate random effects on the estimation of diagnostic accuracy measurements in this scenario.

MODIFIED TEST STATISTICS BY INTER-VOXEL VARIANCE SHRINKAGE WITH AN APPLICATION TO fMRI

Functional Magnetic Resonance Imaging (fMRI) is a non-invasive technique which is commonly used to quantify changes in blood oxygenation and flow coupled to neuronal activation. One of the primary goals of fMRI studies is to identify localized brain regions where neuronal activation levels vary between groups. Single voxel t-tests have been commonly used to determine whether activation related to the protocol differs across groups. Due to the generally limited number of subjects within each study, accurate estimation of variance at each voxel is difficult. Thus, combining information across voxels in the statistical analysis of fMRI data is desirable in order to improve efficiency. Here we construct a hierarchical model and apply an Empirical Bayes framework on the analysis of group fMRI data, employing techniques used in high throughput genomic studies. The key idea is to shrink residual variances by combining information across voxels, and subsequently to construct an improved test statistic in lieu of the classical t-statistic. This hierarchical model results in a shrinkage of voxel-wise residual sample variances towards a common value. The shrunken estimator for voxelspecific variance components on the group analyses outperforms the classical residual error estimator in terms of mean squared error. Moreover, the shrunken test-statistic decreases false positive rate when testing differences in brain contrast maps across a wide range of simulation studies. This methodology was also applied to experimental data regarding a cognitive activation task.

Learning styles vary among general surgery residents: analysis of 12 years of data

PURPOSE: Understanding the learning styles of individuals may assist in the tailoring of an educational program to optimize learning. General surgery faculty and residents have been characterized previously as having a tendency toward particular learning styles. We seek to understand better the learning styles of general surgery residents and differences that may exist within the population. METHODS: The Kolb Learning Style Inventory was administered yearly to general surgery residents at the University of Cincinnati from 1994 to 2006. This tool allows characterization of learning styles into 4 groups: converging, accommodating, assimilating, and diverging. The converging learning style involves education by actively solving problems. The accommodating learning style uses emotion and interpersonal relationships. The assimilating learning style learns by abstract logic. The diverging learning style learns best by observation. Chi-square analysis and analysis of variance were performed to determine significance. RESULTS: Surveys from 1994 to 2006 (91 residents, 325 responses) were analyzed. The prevalent learning style was converging (185, 57%), followed by assimilating (58, 18%), accommodating (44, 14%), and diverging (38, 12%). At the PGY 1 and 2 levels, male and female residents differed in learning style, with the accommodating learning style being relatively more frequent in women and assimilating learning style more frequent in men (Table 1, p < or = 0.001, chi-square test). Interestingly, learning style did not seem to change with advancing PGY level within the program, which suggests that individual learning styles may be constant throughout residency training. If a resident's learning style changed, it tended to be to converging. In addition, no relation exists between learning style and participation in dedicated basic science training or performance on the ABSIT/SBSE. CONCLUSIONS: Our data suggests that learning style differs between male and female general surgery residents but not with PGY level or ABSIT/SBSE performance. A greater understanding of individual learning styles may allow more refinement and tailoring of surgical programs.

Manual pumping test method for characterizing the productivity of drilled wells equipped with rope pumps

Since the introduction of the rope-pump in Nicaragua in the 1990s, the dependence on wells in rural areas has grown steadily. However, little or no attention is paid to rope-pump well performance after installation. Due to financial restraints, groundwater resource monitoring using conventional testing methods is too costly and out of reach of rural municipalities. Nonetheless, there is widespread agreement that without a way to quantify the changes in well performance over time, prioritizing regulatory actions is impossible. A manual pumping test method is presented, which at a fraction of the cost of a conventional pumping test, measures the specific capacity of rope-pump wells. The method requires only sight modifcations to the well and reasonable limitations on well useage prior to testing. The pumping test was performed a minimum of 33 times in three wells over an eight-month period in a small rural community in Chontales, Nicaragua. Data was used to measure seasonal variations in specific well capacity for three rope-pump wells completed in fractured crystalline basalt. Data collected from the tests were analyzed using four methods (equilibrium approximation, time-drawdown during pumping, time-drawdown during recovery, and time-drawdown during late-time recovery) to determine the best data-analyzing method. One conventional pumping test was performed to aid in evaluating the manual method. The equilibrim approximation can be performed while in the field with only a calculator and is the most technologically appropriate method for analyzing data. Results from this method overestimate specific capacity by 41% when compared to results from the conventional pumping test. The other analyes methods, requiring more sophisticated tools and higher-level interpretation skills, yielded results that agree to within 14% (pumping phase), 31% (recovery phase) and 133% (late-time recovery) of the conventional test productivity value. The wide variability in accuracy results principally from difficulties in achieving equilibrated pumping level and casing storage effects in the puping/recovery data. Decreases in well productivity resulting from naturally occuring seasonal water-table drops varied from insignificant in two wells to 80% in the third. Despite practical and theoretical limitations on the method, the collected data may be useful for municipal institutions to track changes in well behavior, eventually developing a database for planning future ground water development projects. Furthermore, the data could improve well-users’ abilities to self regulate well usage without expensive aquifer characterization.

Attempting to improve standardized test results using Study Islands' web-based mastery program

State standardized testing has always been a tool to measure a school’s performance and to help evaluate school curriculum. However, with the school of choice legislation in 1992, the MEAP test became a measuring stick to grade schools by and a major tool in attracting school of choice students. Now, declining enrollment and a state budget struggling to stay out of the red have made school of choice students more important than ever before. MEAP scores have become the deciding factor in some cases. For the past five years, the Hancock Middle School staff has been working hard to improve their students’ MEAP scores in accordance with President Bush's “No Child Left Behind” legislation. In 2005, the school was awarded a grant that enabled staff to work for two years on writing and working towards school goals that were based on the improvement of MEAP scores in writing and math. As part of this effort, the school purchased an internet-based program geared at giving students practice on state content standards. This study examined the results of efforts by Hancock Middle School to help improve student scores in mathematics on the MEAP test through the use of an online program called “Study Island.” In the past, the program was used to remediate students, and as a review with an incentive at the end of the year for students completing a certain number of objectives. It had also been used as a review before upcoming MEAP testing in the fall. All of these methods may have helped a few students perform at an increased level on their standardized test, but the question remained of whether a sustained use of the program in a classroom setting would increase an understanding of concepts and performance on the MEAP for the masses. This study addressed this question. Student MEAP scores and Study Island data from experimental and comparison groups of students were compared to understand how a sustained use of Study Island in the classroom would impact student test scores on the MEAP. In addition, these data were analyzed to determine whether Study Island results provide a good indicator of students’ MEAP performance. The results of the study suggest that there were limited benefits related to sustained use of Study Island and gave some indications about the effectiveness of the mathematics curriculum at Hancock Middle School. These results and implications for instruction are discussed.

Parameter identification of a copper-zeolite SCR catalyst model using reactor data

The selective catalytic reduction system is a well established technology for NOx emissions control in diesel engines. A one dimensional, single channel selective catalytic reduction (SCR) model was previously developed using Oak Ridge National Laboratory (ORNL) generated reactor data for an iron-zeolite catalyst system. Calibration of this model to fit the experimental reactor data collected at ORNL for a copper-zeolite SCR catalyst is presented. Initially a test protocol was developed in order to investigate the different phenomena responsible for the SCR system response. A SCR model with two distinct types of storage sites was used. The calibration process was started with storage capacity calculations for the catalyst sample. Then the chemical kinetics occurring at each segment of the protocol was investigated. The reactions included in this model were adsorption, desorption, standard SCR, fast SCR, slow SCR, NH3 Oxidation, NO oxidation and N2O formation. The reaction rates were identified for each temperature using a time domain optimization approach. Assuming an Arrhenius form of the reaction rates, activation energies and pre-exponential parameters were fit to the reaction rates. The results indicate that the Arrhenius form is appropriate and the reaction scheme used allows the model to fit to the experimental data and also for use in real world engine studies.

Continuous monitoring of bovine spongiform encephalopathy rapid test performance by weak positive tissue controls and quality control charts

Bovine spongiform encephalopathy (BSE) rapid tests and routine BSE-testing laboratories underlie strict regulations for approval. Due to the lack of BSE-positive control samples, however, full assay validation at the level of individual test runs and continuous monitoring of test performance on-site is difficult. Most rapid tests use synthetic prion protein peptides, but it is not known to which extend they reflect the assay performance on field samples, and whether they are sufficient to indicate on-site assay quality problems. To address this question we compared the test scores of the provided kit peptide controls to those of standardized weak BSE-positive tissue samples in individual test runs as well as continuously over time by quality control charts in two widely used BSE rapid tests. Our results reveal only a weak correlation between the weak positive tissue control and the peptide control scores. We identified kit-lot related shifts in the assay performances that were not reflected by the peptide control scores. Vice versa, not all shifts indicated by the peptide control scores indeed reflected a shift in the assay performance. In conclusion these data highlight that the use of the kit peptide controls for continuous quality control purposes may result in unjustified rejection or acceptance of test runs. However, standardized weak positive tissue controls in combination with Shewhart-CUSUM control charts appear to be reliable in continuously monitoring assay performance on-site to identify undesired deviations.