David M. Rocke

Is the good news about compliance good news about cooperation

Recent research on compliance in international regulatory regimes has argued (1) that compliance is generally quite good; (2) that this high level of compliance has been achieved with little attention to enforcement; (3) that those compliance problems that do exist are best addressed as management rather than enforcement problems; and (4) that the management rather than the enforcement approach holds the key to the evolution of future regulatory cooperation in the international system. While the descriptive findings above are largely correct, the policy inferences are dangerously contaminated by endogeneity and selection problems. A high rate of compliance is often the result of states formulating treaties that require them to do little more than they would do in the absence of a treaty. In those cases where noncompliance does occur and where the effects of selection are attenuated, both self-interest and enforcement play significant roles.

Tumor classification by partial least squares using microarray gene expression data

MOTIVATION One important application of gene expression microarray data is classification of samples into categories, such as the type of tumor. The use of microarrays allows simultaneous monitoring of thousands of genes expressions per sample. This ability to measure gene expression en masse has resulted in data with the number of variables p(genes) far exceeding the number of samples N. Standard statistical methodologies in classification and prediction do not work well or even at all when N < p. Modification of existing statistical methodologies or development of new methodologies is needed for the analysis of microarray data. RESULTS We propose a novel analysis procedure for classifying (predicting) human tumor samples based on microarray gene expressions. This procedure involves dimension reduction using Partial Least Squares (PLS) and classification using Logistic Discrimination (LD) and Quadratic Discriminant Analysis (QDA). We compare PLS to the well known dimension reduction method of Principal Components Analysis (PCA). Under many circumstances PLS proves superior; we illustrate a condition when PCA particularly fails to predict well relative to PLS. The proposed methods were applied to five different microarray data sets involving various human tumor samples: (1) normal versus ovarian tumor; (2) Acute Myeloid Leukemia (AML) versus Acute Lymphoblastic Leukemia (ALL); (3) Diffuse Large B-cell Lymphoma (DLBCLL) versus B-cell Chronic Lymphocytic Leukemia (BCLL); (4) normal versus colon tumor; and (5) Non-Small-Cell-Lung-Carcinoma (NSCLC) versus renal samples. Stability of classification results and methods were further assessed by re-randomization studies.

A model for measurement error for gene expression arrays

We introduce a model for measurement error in gene expression arrays as a function of the expression level. This model, together with analysis methods, data transformations, and weighting, allows much more precise comparisons of gene expression, and provides guidance for analysis of background, determination of confidence intervals, and preprocessing data for multivariate analysis.

Examining executive functioning in children with autism spectrum disorder, attention deficit hyperactivity disorder and typical development.

Executive functioning (EF) is an overarching term that refers to neuropsychological processes that enable physical, cognitive, and emotional self-control. Deficits in EF are often present in neurodevelopmental disorders, but examinations of the specificity of EF deficits and direct comparisons across disorders are rare. The current study investigated EF in 7- to 12-year-old children with autism spectrum disorder (ASD), attention deficit hyperactivity disorder (ADHD) and typical development using a comprehensive battery of measures assessing EF, including response inhibition, working memory, cognitive flexibility, planning, fluency and vigilance. The ADHD group exhibited deficits in vigilance, inhibition and working memory relative to the typical group; however, they did not consistently demonstrate problems on the remaining EF measures. Children with ASD showed significant deficits in vigilance compared with the typical group, and significant differences in response inhibition, cognitive flexibility/switching, and working memory compared with both groups. These results lend support for previous findings that show children with autism demonstrate generalized and profound impairment in EF. In addition, the observed deficits in vigilance and inhibitory control suggest that a significant number of children with ASD present with cognitive profiles consistent with ADHD.

Identification of Outliers in Multivariate Data

Abstract New insights are given into why the problem of detecting multivariate outliers can be difficult and why the difficulty increases with the dimension of the data. Significant improvements in methods for detecting outliers are described, and extensive simulation experiments demonstrate that a hybrid method extends the practical boundaries of outlier detection capabilities. Based on simulation results and examples from the literature, the question of what levels of contamination can be detected by this algorithm as a function of dimension, computation time, sample size, contamination fraction, and distance of the contamination from the main body of data is investigated. Software to implement the methods is available from the authors and STATLIB.

Multi-class cancer classification via partial least squares with gene expression profiles

MOTIVATION Discrimination between two classes such as normal and cancer samples and between two types of cancers based on gene expression profiles is an important problem which has practical implications as well as the potential to further our understanding of gene expression of various cancer cells. Classification or discrimination of more than two groups or classes (multi-class) is also needed. The need for multi-class discrimination methodologies is apparent in many microarray experiments where various cancer types are considered simultaneously. RESULTS Thus, in this paper we present the extension to the classification methodology proposed earlier Nguyen and Rocke (2002b; Bioinformatics, 18, 39-50) to classify cancer samples from multiple classes. The methodologies proposed in this paper are applied to four gene expression data sets with multiple classes: (a) a hereditary breast cancer data set with (1) BRCA1-mutation, (2) BRCA2-mutation and (3) sporadic breast cancer samples, (b) an acute leukemia data set with (1) acute myeloid leukemia (AML), (2) T-cell acute lymphoblastic leukemia (T-ALL) and (3) B-cell acute lymphoblastic leukemia (B-ALL) samples, (c) a lymphoma data set with (1) diffuse large B-cell lymphoma (DLBCL), (2) B-cell chronic lymphocytic leukemia (BCLL) and (3) follicular lymphoma (FL) samples, and (d) the NCI60 data set with cell lines derived from cancers of various sites of origin. In addition, we evaluated the classification algorithms and examined the variability of the error rates using simulations based on randomization of the real data sets. We note that there are other methods for addressing multi-class prediction recently and our approach is along the line of Nguyen and Rocke (2002b; Bioinformatics, 18, 39-50). CONTACT dnguyen@stat.tamu.edu; dmrocke@ucdavis.edu

Predicting ligand binding to proteins by affinity fingerprinting

BACKGROUND There are many ways to represent a molecules properties, including atomic-connectivity drawings, NMR spectra, and molecular orbital models. Prior methods for predicting the biological activity of compounds have largely depended on these physical representations. Measuring a compounds binding potency against a small reference panel of diverse proteins defines a very different representation of the molecule, which we call an affinity fingerprint. Statistical analysis of such fingerprints provides new insights into aspects of binding interactions that are shared among a wide variety of proteins. These analyses facilitate prediction of the binding properties of these compounds assayed against new proteins. RESULTS Affinity fingerprints are reported for 122 structurally-diverse compounds using a reference panel of eight proteins that collectively are able to generate unique fingerprints for about 75% of the small organic compounds tested. Application of multivariate regression techniques to this database enables the creation of computational surrogates to represent new proteins that are surprisingly effective at predicting binding potencies. We illustrate this for two enzymes with no previously recognizable similarity to each other or to any of the reference proteins. Fitting of analogous computational surrogates to four other proteins confirms the generality of the method; when applied to a fingerprinted library of 5000 compounds, several sub-micromolar hits were correctly predicted. CONCLUSIONS An affinity fingerprint database, which provides a rich source of data defining operational similarities among proteins, can be used to test theories of cryptic homology unexpected from current understanding of protein structure. Practical applications to drug design include efficient pre-screening of large numbers of compounds against target proteins using fingerprint similarities, supplemented by a small number of empirical measurements, to select promising compounds for further study.

A Two-Component Model for Measurement Error in Analytical Chemistry

In this article, we propose and test a new model for measurement error in analytical chemistry. Often, the standard deviation of analytical errors is assumed to increase proportionally to the concentration of the analyte, a model that cannot be used for very low concentrations. For near-zero amounts, the standard deviation is often assumed constant, which does not apply to larger quantities. Neither model applies across the full range of concentrations of an analyte. By positing two error components, one additive and one multiplicative, we obtain a model that exhibits sensible behavior at both low and high concentration levels. We use maximum likelihood estimation and apply the technique to toluene by gas-chromatography/mass-spectrometry and cadmium by atomic absorption spectroscopy.

Partial least squares proportional hazard regression for application to DNA microarray survival data

MOTIVATION Microarrays are increasingly used in cancer research. When gene transcription data from microarray experiments also contains patient survival information, it is often of interest to predict the survival times based on the gene expression. In this paper we consider the well-known proportional hazard (PH) regression model for survival analysis. Ordinarily, the PH model is used with a few covariates and many observations (subjects). We consider here the case that the number of covariates, p, exceeds the number of samples, N, a setting typical of gene expression data from DNA microarrays. RESULTS For a given vector of response values which are survival times and p gene expressions (covariates) we examine the problem of how to predict the survival probabilities, when N << p. The approach taken to cope with the high dimensionality is to reduce the dimension using partial least squares with the response variable as the vector of survival times. After dimension reduction, the extracted PLS gene components are then used as covariates in a PH regression to predict the survival probabilities. We demonstrate the use of the methodology on two cDNA gene expression data sets, both containing survival data. The first data set contains 40 diffuse large B-cell lymphoma (DLBCL) tissue samples and the second data set contains 49 tissue samples from patients with locally advanced breast cancer in a prospective study.

IATROGENIC HARM CAUSED BY DIAGNOSTIC ERRORS IN FIBRODYSPLASIA OSSIFICANS PROGRESSIVA

Background. Little is known about diagnostic errors for a disease worldwide. Such errors could alter the diseases natural history, especially if unwarranted interventions cause irreversible harm. Fibrodysplasia ossificans progressiva (FOP), a rare, autosomal dominant genetic disease characterized by episodes of permanent heterotopic ossification of soft tissues, occurs worldwide without racial, ethnic, or geographic predilection. There is no effective treatment, and soft-tissue trauma (eg, biopsies, surgical procedures, intramuscular injections, or mandibular blocks for dental procedures) and viral illnesses are likely to induce episodes of rapidly progressive heterotopic ossification, with resultant permanent loss of motion in the affected area. Accurate diagnoses can be made on the basis of the clinical findings of tumor-like swellings on the head, neck, back, or shoulders and characteristic short great toes with hallux valgus-like malformations and missing interphalangeal joints. On the basis of conversations with numerous individuals with FOP, we suspected that diagnostic errors with FOP are common and often associated with inappropriate and harmful diagnostic and therapeutic procedures. Objective. To document the frequency of diagnostic errors with FOP and complications resulting from misdiagnoses. Design. A questionnaire requesting detailed demographic, diagnostic, and treatment information was sent to all 269 patient-members of the International FOP Association; the sampling frame included >90% of all known FOP patients worldwide. We received 138 replies (51% response) from 25 countries. The age range was 2 to 71 years; there were 78 female subjects and 60 male subjects. In addition, to assess the availability and adequacy of information about FOP, we reviewed 184 English-language textbooks in relevant specialties published in the past 20 years. Results. Incorrect diagnoses were given initially to 87% of individuals with FOP. This astonishing rate of diagnostic errors occurred worldwide, regardless of ethnicity, geographic background, or misdiagnosing physicians specialty. The most common incorrect diagnosis was cancer (32%). The mean period from the onset of symptoms to correct diagnosis was 4.1 years, and the median number of physicians consulted before the correct diagnosis of FOP was 6. For 67% of patients, unnecessary invasive procedures (biopsies) were performed; 68% received inappropriate therapies. Forty-nine percent of all patients reported permanent loss of mobility resulting from invasive medical interventions that caused posttraumatic ossification. Notably, only 8% of the 184 textbooks that were reviewed contained adequate descriptions of FOP, including the caution that trauma can accelerate the process of heterotopic ossification. Conclusions. Diagnostic errors and inappropriate medical procedures, which may lead to permanent harm, can alter the natural history of a disease. In FOP, the astonishing rates of diagnostic errors and inappropriate invasive medical procedures likely result from lack of physician awareness because of failure of information transfer.