Selection models and pattern-mixture models for incomplete data with covariates

Most models for incomplete data are formulated within the selection model framework. This paper studies similarities and differences of modeling incomplete data within both selection and pattern-mixture settings. The focus is on missing at random mechanisms and on categorical data. Point and interval estimation is discussed. A comparison of both approaches is done on side effects in a psychiatric study.     

The Effect of Cauda Equina Constriction on Nitric Oxide Synthase Activity

Nitric oxide synthase (NOS) activity was studied in the gray and white matter regions of the spinal cord 2 and 5 days after multiple cauda equina constrictions of the central processes of L7-Co5 dorsal root ganglia neurons. The results show considerable differences in enzyme activity in the thoracic, upper lumbar, lower lumbar, and sacral segments. Increased NOS activity was observed at 2 days after multiple cauda equina constrictions in the dorsal, lateral, and ventral columns of the lower lumbar segments and in the ventral column of the upper lumbar segments. The values returned to control levels within 5 postconstriction days. In the lateral columns of thoracic segments taken 2 and 5 days after surgery, NOS activity was enhanced by 54% and 55% and in the upper lumbar segments by 130% and 163%, respectively. Multiple cauda equina constrictions performed surgically for 2 and 5 days caused a significant increase in NOS activity predominantly in the gray matter regions of thoracic segments. A quite different response was found 5 days postconstriction in the upper lumbar segments, where the enzyme activity was significantly decreased in the dorsal horn, intermediate zone, and ventral horn. No such extreme differences could be seen in the lower lumbar segments, where NOS activity was significantly enhanced only in the ventral horn. The data correspond with a higher number of NOS immunoreactive somata, quantitatively evaluated in the ventral horn of the lower lumbar segments at 5 days after multiple cauda equina constrictions. While the great region-dependent heterogeneity in NOS activity seen 2 and 5 days after multiple cauda equina constrictions is quite apparent and suggestive of an active role played by nitric oxide in neuroprotective or neurotoxic processes occurring in the gray and white matter of the spinal cord, the extent of damage or the degree of neuroprotection caused by nitric oxide in compartmentalized gray and white matter in this experimental paradigm would be possible only using longer postconstriction periods.     

Molecular phylogeny of the hexagrammid fishes using a multi-locus approach

Ideally, organisms are grouped into monophyletic assemblages reflecting their evolutionary histories. Single (molecular) markers can reflect the evolutionary history of the marker, rather than the species in question, therefore, phylogenetic relationships should be inferred from adequate sampling of characters. Because the use of multiple loci greatly improves the resolving power of the molecular assay, we constructed a molecular phylogeny of the family Hexagrammidae based on six loci, including two mitochondrial and four nuclear loci. The resulting molecular phylogeny, from the combined data, was significantly different from the morphological topology suggested by Shinohara [Memoirs of the Faculty of Fisheries, Hokkaido University 41 (1994) 1]. Our data support a monophyletic assemblage for the genera Hexagrammos and Pleurogrammus. However, other taxa traditionally included in the family Hexagrammidae did not form a monophyletic assemblage. The monotypic genus Ophiodon was more closely associated with cottids than with other hexagrammids. Our data concur with the morphological topology in that the genera Zaniolepis and Oxylebius formed a monophyletic clade, which was distinct and basal to the remaining hexagrammids, seven cottids and one agonid.     

Disruption of thiol homeostasis and nitrosative stress in the cerebrospinal fluid of patients with active multiple sclerosis: evidence for a protective role of acetylcarnitine

Recent studies suggest that NO and its reactive derivative peroxynitrite are implicated in the pathogenesis of multiple sclerosis (MS). Patients dying with MS demonstrate increased astrocytic inducible nitric oxide synthase activity, as well as increased levels of iNOS mRNA. Peroxynitrite is a strong oxidant capable of damaging target tissues, particularly the brain, which is known to be endowed with poor antioxidant buffering capacity. Inducible nitric oxide synthase is upregulated in the central nervous system (CNS) of animals with experimental allergic encephalomyelitis (EAE) and in patients with MS. We have recently demonstrated in patients with active MS a significant increase of NOS activity associated with increased nitration of proteins in the cerebrospinal fluid (CSF). Acetylcarnitine is proposed as a therapeutic agent for several neurodegenerative disorders. Accordingly, in the present study, MS patients were treated for 6 months with acetylcarnitine and compared with untreated MS subjects or with patients noninflammatory neurological conditions, taken as controls. Western blot analysis showed in MS patients increased nitrosative stress associated with a significant decrease of reduced glutathione (GSH). Increased levels of oxidized glutathione (GSSG) and nitrosothiols were also observed. Interestingly, treatment of MS patients with acetylcarnitine resulted in decreased CSF levels of NO reactive metabolites and protein nitration, as well as increased content of GSH and GSH/GSSG ratio. Our data sustain the hypothesis that nitrosative stress is a major consequence of NO produced in MS-affected CNS and implicate a possible important role for acetylcarnitine in protecting brain against nitrosative stress, which may underlie the pathogenesis of MS.     

Roles of conserved methionine residues in tobacco acetolactate synthase

Acetolactate synthase (ALS) catalyzes the first common step in the biosynthesis of valine, leucine, and isoleucine. ALS is the target of several classes of herbicides, including the sulfonylureas, the imidazolinones, and the triazolopyrimidines. The conserved methionine residues of ALS from plants were identified by multiple sequence alignment using ClustalW. The alignment of 17 ALS sequences from plants revealed 149 identical residues, seven of which were methionine residues. The roles of three well-conserved methionine residues (M350, M512, and M569) in tobacco ALS were determined using site-directed mutagenesis. The mutation of M350V, M512V, and M569V inactivated the enzyme and abolished the binding affinity for cofactor FAD. Nevertheless, the secondary structure of each of the mutants determined by CD spectrum was not affected significantly by the mutation. Both M350C and M569C mutants were strongly resistant to three classes of herbicides, Londax (a sulfonylurea), Cadre (an imidazolinone), and TP (a triazolopyrimidine), while M512C mutant did not show a significant resistance to the herbicides. The mutant M350C was more sensitive to pH change, while the mutant M569C showed a profile for pH dependence activity similar to that of wild type. These results suggest that M512 residue is likely located at or near the active site, and that M350 and M569 residues are probably located at the overlapping region between the active site and a common herbicide binding site.     

Bayesian inferences in the Cox model for order-restricted hypotheses

In studying the relationship between an ordered categorical predictor and an event time, it is standard practice to include dichotomous indicators of the different levels of the predictor in a Cox model. One can then use a multiple degree-of-freedom score or partial likelihood ratio test for hypothesis testing. Often, interest focuses on comparing the null hypothesis of no difference to an order-restricted alternative, such as a monotone increase across levels of a predictor. This article proposes a Bayesian approach for addressing hypotheses of this type. We reparameterize the Cox model in terms of a cumulative product of parameters having conjugate prior densities, consisting of mixtures of point masses at one, and truncated gamma densities. Due to the structure of the model, posterior computation can proceed via a simple and efficient Gibbs sampling algorithm. Posterior probabilities for the global null hypothesis and subhypotheses, comparing the hazards for specific groups, can be calculated directly from the output of a single Gibbs chain. The approach allows for level sets across which a predictor has no effect. Generalizations to multiple predictors are described, and the method is applied to a study of emergency medical treatment for stroke.     

A local influence approach applied to binary data from a psychiatric study

Recently, a lot of concern has been raised about assumptions needed in order to fit statistical models to incomplete multivariate and longitudinal data. In response, research efforts are being devoted to the development of tools that assess the sensitivity of such models to often strong but always, at least in part, unverifiable assumptions. Many efforts have been devoted to longitudinal data, primarily in the selection model context, although some researchers have expressed interest in the pattern-mixture setting as well. A promising tool, proposed by Verbeke et al. (2001, Biometrics 57, 43-50), is based on local influence (Cook, 1986, Journal of the Royal Statistical Society, Series B 48, 133-169). These authors considered the Diggle and Kenward (1994, Applied Statistics 43, 49-93) model, which is based on a selection model, integrating a linear mixed model for continuous outcomes with logistic regression for dropout. In this article, we show that a similar idea can be developed for multivariate and longitudinal binary data, subject to nonmonotone missingness. We focus on the model proposed by Baker, Rosenberger, and DerSimonian (1992, Statistics in Medicine 11, 643-657). The original model is first extended to allow for (possibly continuous) covariates, whereafter a local influence strategy is developed to support the model-building process. The model is able to deal with nonmonotone missingness but has some limitations as well, stemming from the conditional nature of the model parameters. Some analytical insight is provided into the behavior of the local influence graphs.     

Binary culture biofilm formation by Stenotrophomonas maltophilia and Fusarium oxysporum

Binary culture biofilm formation by Stenotrophomonas maltophilia and Fusarium oxysporum was investigated using the recirculating modified Robbins device batch culture system. Sequential attachment studies were carried out in the Robbins device on PVC and glass surfaces, with each species as either the first or the second colonizer. Different surfaces had no significant effect on total numbers of S. maltophilia and F. oxysporum in the binary population biofilm. The attachment of the second colonizer was not influenced significantly by the previous attachment of the first colonizer. These results were confirmed using scanning electron micrographs. Journal of Industrial Microbiology & Biotechnology (2001) 26, 178–183. Received 06 July 1999/ Accepted in revised form 02 November 2000     

The fixation probability of a beneficial allele in a population dividing by binary fission

We derive formulae for the fixation probability, P, of a rare benefical allele segregating in a population of fixed size which reproduces by binary fission, in terms of the selection coefficient for the beneficial allele, s. We find that an earlier result P 4s does not depend on the assumption of binary fission, but depends on an assumption about the ordering of events in the life cycle. We find that P 2s for mutations occurring during chromosome replication and P 2.8s for mutations occurring at random times between replication events.