Monitoring complex media fermentations with near-infrared spectroscopy: comparison of different variable selection methods

Near-infrared spectroscopy (NIRS) is known to be a suitable technique for rapid fermentation monitoring. Industrial fermentation media are complex, both chemically (ill-defined composition) and physically (multiphase sample matrix), which poses an additional challenge to the development of robust NIRS calibration models. We investigated the use of NIRS for at-line monitoring of the concentration of clavulanic acid during an industrial fermentation. An industrial strain of Streptomyces clavuligerus was cultivated at 200-L scale for the production of clavulanic acid. Partial least squares (PLS) regression was used to develop calibration models between spectral and analytical data. In this work, two different variable selection methods, genetic algorithms (GA) and PLS-bootstrap, were studied and compared with models built using all the spectral variables. Calibration models for clavulanic acid concentration performed well both on internal and external validation. The two variable selection methods improved the predictive ability of the models up to 20%, relative to the calibration model built using the whole spectra.     

Estimation in Bayesian disease mapping

Recent work on Bayesian inference of disease mapping models discusses the advantages of the fully Bayesian (FB) approach over its empirical Bayes (EB) counterpart, suggesting that FB posterior standard deviations of small-area relative risks are more reflective of the uncertainty associated with the relative risk estimation than counterparts based on EB inference, since the latter fail to account for the variability in the estimation of the hyperparameters. In this article, an EB bootstrap methodology for relative risk inference with accurate parametric EB confidence intervals is developed, illustrated, and contrasted with the hyperprior Bayes. We elucidate the close connection between the EB bootstrap methodology and hyperprior Bayes, present a comparison between FB inference via hybrid Markov chain Monte Carlo and EB inference via penalized quasi-likelihood, and illustrate the ability of parametric bootstrap procedures to adjust for the undercoverage in the "naive" EB interval estimates. We discuss the important roles that FB and EB methods play in risk inference, map interpretation, and real-life applications. The work is motivated by a recent analysis of small-area infant mortality rates in the province of British Columbia in Canada.     

On studentising and blocklength selection for the bootstrap on time series

For independent data, non-parametric bootstrap is realised by resampling the data with replacement. This approach fails for dependent data such as time series. If the data generating process is at least stationary and mixing, the blockwise bootstrap by drawing subsamples or blocks of the data saves the concept. For the blockwise bootstrap a blocklength has to be selected. We propose a method for selecting the optimal blocklength. To improve the finite size properties of the blockwise bootstrap, studentised statistics is considered. If the statistic can be represented as a smooth function model this studentisation can be approximated efficiently. The studentised blockwise bootstrap method is applied for testing hypotheses on medical time series.     

Dental relatedness corresponding to mortuary patterning at Huaca Loro, Peru

Within and between tombs at the 1,000-year-old site of Huaca Loro on north coastal Peru, interment characteristics vary to an intriguing degree. Following and elaborating upon prior intracemetery studies, biological relatedness among associated burial groupings was assessed using 23 dental characters (assuming familial allele segregation) for 29 individuals. Biological patterning was based on multivariate distance between individuals using all traits, rather than the previously widespread reliance upon univariate comparison of each trait separately within samples. This multivariate approach did seem more informative. Statistically significant variation of biological similarities and dissimilarities corresponded to spatial groupings and also to various specific archaeological indications of the cohesiveness, or lack thereof, of interment pattern. The partition of biological distances among tombs at Huaca Loro supports the archaeological evidence that the tombs represent a planned elite cemetery.     

The surface area of monomeric proteins: significance of power law behavior

The coefficients in a power law fit of accessible area versus molecular weight for high-resolution monomeric protein structures are assessed with respect to statistical accuracy using bootstrap analyses, and with respect to physical significance using model systems and the concept of roughness or fractal structure of the protein surface.     

Odontometric differences betweenCercopithecus aethiops populations: A simulation approach to estimating confidence limits for Penrose’s shape coefficient

A modified bootstrapping procedure is described by means of which standard errors and confidence limits may be determined for the Penrose shape coefficient. This method is then applied to odontometric data derived from four closely related groups of primates: Cercopithecus aethiops pygerythrus, C. a. sabaeus,St. Kitt’s green monkey (derived from C. a. sabaeus),and C. a. centralis.Although statistically significant distances were found to exist between these groups, one shape coefficient was significantly greater than the others: that between C. a. pygerythrusand C. a. centralis.Low levels of sexual dimorphism characterized the shape coefficients, with distances based on mandibular teeth being greater than those derived from maxillary teeth.     

Molecular Phylogenetics at the Felsenstein Zone: Approaching the Strepsiptera Problem Using 5.8S and 28S rDNA Sequences

Recent efforts to reconstruct the phylogenetic position of the insect order Strepsiptera have elicited a major controversy in molecular phylogenetics. We sequenced the 5.8S rDNA and major parts of the 28S rDNA 5′ region of the strepsipteran speciesStylops melittae.Their evolutionary dynamics were analyzed together with previously published insect rDNA sequences to identify tree estimation bias risks and to explore additional sources of phylogenetic information. Several major secondary structure changes were found as being autapomorphic for the Diptera, the Strepsiptera, or the Archaeognatha. Besides elevated substitution rates a significant AT bias was present in dipteran and strepsipteran 28S rDNA which, however, was restricted to stem sites in the Diptera while also affecting single-stranded sites in the Strepsiptera. When dipteran taxa were excluded from tree estimation all methods consistently supported the placement of Strepsiptera to within the Holometabola. When dipteran taxa were included maximum likelihood continued to favor a sister-group relationship of Strepsiptera with Mecoptera while remaining methods strongly supported a sister-group relationship with Diptera. Parametric bootstrap analysis revealed maximum likelihood as a consistent estimator if rate heterogeneity across sites was taken into account. Though the position of Strepsiptera within Holometabola remains elusive, we conclude that the evolution of dipteran and strepsipteran rDNA involved similar yet independent changes of substitution parameters.     

SEASONAL PATTERNS IN THE ABUNDANCE AND DISTRIBUTION OF CALIFORNIA CETACEANS, 1991–1992

This study presents a detailed seasonal comparison of the abundance and distribution of cetaceans within 100-150 nmi (185-278 km) of the California coast during 1991 and 1992. The results of a shipboard line-transect survey conducted in July-November 1991 ("summer") were compared to those from aerial line-transect surveys conducted in March-April 1991 and February-April 1992 ("winter"). Using a confidence-interval-based bootstrap procedure, abundance estimates for six of the eleven species included in the comparison exhibited significant (α= 0.05) differences between the winter and summer surveys. Pacific white-sided dolphins ( Lagenorhynchus obliquidens ), Risso's dolphins ( Grampus griseus ), common dolphins ( Delphinus spp.), and northern right whale dolphins ( Lissodelphis borealis ) were significantly more abundant in winter. The abundance of blue whales ( Balaenoptera musculuss ) and gray whales ( Eschrichtius robustus ) reflected well-documented migratory patterns. Fin whales ( B. physalus ) were significantly more abundant during summer. No significant differences in seasonal abundance were identified for Dall's porpoises ( Phocoenoides dalli ), bottlenose dolphins ( Tursiops truncatus ), killer whales ( Orcinus orca ), sperm whales ( Physeter macrocephalus ), or humpback whales ( Megaptera novaeangliae ). Significant north/south shifts in distribution were found for Dall's porpoises, common dolphins, and Pacific white-sided dolphins, and significant inshore/offshore differences were identified for northern right whale dolphins and humpback whales.