Likelihood and conditional likelihood inference for generalized additive mixed models for clustered data

Lin and Zhang (J. Roy. Statist. Soc. Ser. B 61 (1999) 381) proposed the generalized additive mixed model (GAMM) as a framework for analysis of correlated data, where normally distributed random effects are used to account for correlation in the data, and proposed to use double penalized quasi-likelihood (DPQL) to estimate the nonparametric functions in the model and marginal likelihood to estimate the smoothing parameters and variance components simultaneously. However, the normal distributional assumption for the random effects may not be realistic in many applications, and it is unclear how violation of this assumption affects ensuing inferences for GAMMs. For a particular class of GAMMs, we propose a conditional estimation procedure built on a conditional likelihood for the response given a sufficient statistic for the random effect, treating the random effect as a nuisance parameter, which thus should be robust to its distribution. In extensive simulation studies, we assess performance of this estimator under a range of conditions and use it as a basis for comparison to DPQL to evaluate the impact of violation of the normality assumption. The procedure is illustrated with application to data from the Multicenter AIDS Cohort Study (MACS).     

Identification,synthesis and conformational study of a CN(R,S) synthon by-product: (5R, 10R)-5,10-diphenyl-1,6-diaza-3,8-dioxabicyclo[4.4.1]undecane

The title compound, containing a new heterocyclic skeleton, was identified by X-ray crystallography as the product of condensation of (R)-phenylglycinol with an excess of formaldehyde. The molecule adopts a rigid double twist-chair conformation in both solid and solution states.     

A short proof for a generalization of Vizing's theorem

For a simple graph of maximum degree Δ, it is always possible to color the edges with Δ + 1 colors (Vizing); furthermore, if the set of vertices of maximum degree is independent, Δ colors suffice (Fournier). In this article, we give a short constructive proof of an extension of these results to multigraphs. Instead of considering several color interchanges along alternating chains (Vizing, Gupta), using counting arguments (Ehrenfeucht, Faber, Kierstead), or improving nonvalid colorings with Fournier's Lemma, the method of proof consists of using one single easy transformation, called “sequential recoloring”, to augment a partial k-coloring of the edges.     

The silicon effect on the regioselectivity of the Tsuji-Trost reaction. Experimental and theoretical approaches

Regioselectivity of the Tsuji-Trost reaction on allyl acetates and carbonates substituted with silyl groups at the olefinic moiety has been analyzed. Silicon atom in the central carbon atom increases the stability of the intermediate π-allylpalladium cation with respect to the isomeric π-allylpalladium cation featuring the silicon in a terminal carbon atom.     

Du discret au continu pour des modèles de réseaux aléatoires d'atomes

We study in this Note the continuum (macroscopic) limit for some atomistic models for crystals. The purpose is to derive densities of mechanical energies from microscopic models. In contrast to the setting of a previous study, where the microscopic structure was assumed to be periodic, it is modelled here by a stochastic lattice, which enjoys some stationarity and ergodicity properties, following notions previously introduced elsewhere. To cite this article: X. Blanc et al., C. R. Acad. Sci. Paris, Ser. I 342 (2006).     

Quaternionic-Kähler manifolds and conformal geometry

Research funded in part by NSF grant DMS-8704401     

Species differences in the metabolism of arsenic compounds

Humans are exposed via air, water and food to a number of different arsenic compounds, the physical, chemical, and toxicological properties of which may vary considerably. In people eating much fish and shellfish the intake of organic arsenic compounds, mainly arsenobetaine, may exceed 1000 μg As per day, while the average daily intake of inorganic arsenic is in the order of 10–20 μg in most countries. Arsenobetaine, and most other arsenic compounds in food of marine origin, e.g. arsenocholine, trimethylarsine oxide and methylarsenic acids, are rapidly excreted in the urine and there seem to be only minor differences in metabolism between animal species. Trivalent inorganic arsenic (AsIII) is the main form of arsenic interacting with tissue constituents, due to its strong affinity for sulfhydryl groups. However, a substantial part of the absorbed AsIII is methylated in the body to less reactive metabolities, methylarsonic acid (MMA) and dimethylarsinic acid (DMA), which are rapidly excreted in the urine. All the different steps in the arsenic biotransformation in mammals have not yet been elucidated, but it seems likely that the methylation takes place mainly in the liver by transfer of methyl groups from S-adenosylmethionine to arsenic in its trivalent oxidation state. A substantial part of absorbed arsenate (AsV) is reduced to AsIII before being methylated in the liver. There are marked species differences in the methylation of inorganic arsenic. In most animal species DMA is the main metabolite. Compared with human subjects, very little MMA is produced. The marmoset monkey is the only species which has been shown unable to methylate inorganic arsenic. In contrast to other species, the rat shows a marked binding of DMA to the hemoglobin, which results in a low rate of urinary excretion of arsenic.     

Unexpected end‐groups of poly(acrylic acid) prepared by RAFT polymerization

Low‐molecular‐weight poly(acrylic acid) (PAA) was synthesized by reversible addition fragmentation chain transfer polymerization with a trithiocarbonate as chain‐transfer agent (CTA). With a combination of NMR spectroscopy and matrix‐assisted laser desorption/ionization time‐of‐flight mass spectrometry, the PAA end‐groups of the polymer were analyzed before and after neutralization by sodium hydroxide. The polymer prior to neutralization is made up of the expected trithiocarbonate chain‐ends and of the H‐terminated chains issued from a reaction of transfer to solvent. After neutralization, the trithiocarbonates are transformed into thiols, disulfides, thiolactones, and additional H‐terminated chains. By quantifying the different end‐groups, it was possible to demonstrate that fragmentation is the rate limiting step in the transfer reaction. © 2004 Wiley Periodicals, Inc. J Polym Sci Part A: Polym Chem 42: 5439–5462, 2004