Multiple imputation for missing continuous and dichotomous data: The effects of patterns of missingness and variable correlations on type I error with small samples

The purpose of this study is to investigate the effects of predictor variable correlations and patterns of missingness with dichotomous and/or continuous data in small samples when missing data is multiply imputed. Missing data of predictor variables is multiply imputed under three different multivariate models: the multivariate normal model for continuous data, the multinomial model for dichotomous data and the general location model for mixed dichotomous and continuous data. Subsequent to the multiple imputation process, Type I error rates of the regression coefficients obtained with logistic regression analysis are estimated under various conditions of correlation structure, sample size, type of data and patterns of missing data. The distributional properties of average mean, variance and correlations among the predictor variables are assessed after the multiple imputation process. ^ For continuous predictor data under the multivariate normal model, Type I error rates are generally within the nominal values with samples of size n = 100. Smaller samples of size n = 50 resulted in more conservative estimates (i.e., lower than the nominal value). Correlation and variance estimates of the original data are retained after multiple imputation with less than 50% missing continuous predictor data. For dichotomous predictor data under the multinomial model, Type I error rates are generally conservative, which in part is due to the sparseness of the data. The correlation structure for the predictor variables is not well retained on multiply-imputed data from small samples with more than 50% missing data with this model. For mixed continuous and dichotomous predictor data, the results are similar to those found under the multivariate normal model for continuous data and under the multinomial model for dichotomous data. With all data types, a fully-observed variable included with variables subject to missingness in the multiple imputation process and subsequent statistical analysis provided liberal (larger than nominal values) Type I error rates under a specific pattern of missing data. It is suggested that future studies focus on the effects of multiple imputation in multivariate settings with more realistic data characteristics and a variety of multivariate analyses, assessing both Type I error and power. ^

Contents of organic compounds in sediments from DSDP Legs 1 and 4

Fifteen sediment samples were studied from five drill sites recovered by the Glomar Challenger on Legs I and IV in the Gulf of Mexico and western Atlantic. This study concentrated on compounds derived from biogenic precursors, namely: (1) hydrocarbons, (2) fatty acids, (3) pigments and (4) amino acids. Carbon isotope (dC13) data [values <(-26)?, relative to PDB], long-chain n-alkyl hydrocarbons (>>C27) with odd carbon numbered molecules dominating even carbon numbered species, and presence of perylene proved useful as possible indicators for terrigenous contributions to the organic matter in some samples. Apparently land-derived organic matter can be transported for distances over 1000 km into the ocean and their source still recognized. The study was primarily designed to investigate: (i) the sources of the organic matter present in the sediment, (ii) their stability with time of accumulation and (iii) the conditions necessary for in situ formation of new compounds.

Chemical and isotope compositions of main phases of metalliferous sediments from holes of DSDP Leg 92

The carbonate-free metalliferous fraction of thirty-nine sediment samples from four DSDP Leg 92 sites has been analyzed for 12 elements, and a subset of 16 samples analyzed for Pb isotopic composition. The main geochemical features of this component are as follows: i) very high concentrations of Fe and Mn, typically 25-39% and 5-14%, respectively; ii) Al and Ca contents generally less than 2% and 5%, respectively; iii) high Cu (1000-2000 ppm), and Zn and Ni (500-1000 ppm) values; and iv) Co and Pb concentrations of 100-250 ppm. In terms of element partitioning within the metalliferous fraction, amorphous to poorly crystallized oxide-oxyhydroxides removed by the second leach carry virtually all of the Mn, and about 90% of the Ca, Sr and Ni. The well-crystallized goethite-rich material removed by the third leach carries the majority of Fe, Cu, and Pb. These relations hold for sediments as young as ~1-2 Ma, indicating early partitioning of hydrothermal Fe and Mn into separate phases. Calculated mass accumulation rates (MAR) for Fe, Mn, Cu, Pb, Zn and Ni in the bulk sediment show the same overall trends at three of the sites, with greatest MAR values near the basement, and a general decrease in MAR values towards the tops of the holes (for sediments deposited above the lysocline). These relations strongly support the concept of a declining hydrothermal contribution of these elements away from a ridge axis. Nevertheless, MAR values for these metals up to ~200 km from the ridge axis are orders of magnitude higher than on abyssal seafloor plains where there is no hydrothermal influence. Mn/Fe ratios throughout the sediment column at two sites indicate that the composition of the hydrothermal precipitates changed during transport through seawater, becoming significantly depleted in Mn beyond ~200-300 km from the axis, but maintaining roughly the same proportion of Fe. Most of the Pb isotope data for the Leg 92 metalliferous sediments form approximately linear arrays in the conventional isotopic plots, extending from the middle of the field for mid-ocean ridge basalts toward the field for Mn nodules. The array of data lying between these two end-members is most readily interpreted in terms of simple linear mixing of Pb derived from basaltic and seawater end-member sources. The least radiogenic sediments reflect the average Pb isotope composition of discharging hydrothermal solutions and ocean-ridge basalt at the EPR over the ~4-8 Ma B.P. interval. Pb in sediments deposited up to 250 km from the axis can be almost entirely of basaltic-hydrothermal origin. Lateral transport of some basaltic Pb by ocean currents appears to extend to distances of at least 1000 km west of the East Pacific Rise.

C and H geochemistry of altered oceanic crus of ODP/IODP Hole 1256D

Carbon and hydrogen concentrations and isotopic compositions were measured in 19 samples from altered oceanic crust cored in ODP/IODP Hole 1256D through lavas, dikes down to the gabbroic rocks. Bulk water content varies from 0.32 to 2.14 wt% with dD values from -64per mil to -25per mil. All samples are enriched in water relative to fresh basalts. The dD values are interpreted in terms of mixing between magmatic water and another source that can be either secondary hydrous minerals and/or H contained in organic compounds such as hydrocarbons. Total CO2, extracted by step-heating technique, ranges between 564 and 2823 ppm with d13C values from -14.9per mil to -26.6per mil. As for water, these altered samples are enriched in carbon relative to fresh basalts. The carbon isotope compositions are interpreted in terms of a mixing between two components: (1) a carbonate with d13C = -4.5per mil and (2) an organic compound with d13C = -26.6per mil. A mixing model calculation indicates that, for most samples (17 of 19), more than 75% of the total C occurs as organic compounds while carbonates represent less than 25%. This result is also supported by independent estimates of carbonate content from CO2 yield after H3PO4 attack. A comparison between the carbon concentration in our samples, seawater DIC (Dissolved Inorganic Carbon) and DOC (Dissolved Organic Carbon), and hydrothermal fluids suggests that CO2 degassed from magmatic reservoirs is the main source of organic C addition to the crust during the alteration process. A reduction step of dissolved CO2 is thus required, and can be either biologically mediated or not. Abiotic processes are necessary for the deeper part of the crust (>1000 mbsf) because alteration temperatures are greater than any hyperthermophilic living organism (i.e. T > 110 °C). Even if not required, we cannot rule out the contribution of microbial activity in the low-temperature alteration zones. We propose a two-step model for carbon cycling during crustal alteration: (1) when "fresh" oceanic crust forms at or close to ridge axis, alteration starts with hot hydrothermal fluids enriched in magmatic CO2, leading to the formation of organic compounds during Fischer-Tropsch-type reactions; (2) when the crust moves away from the ridge axis, these interactions with hot hydrothermal fluids decrease and are replaced by seawater interactions with carbonate precipitation in fractures. Taking into account this organic carbon, we estimate C isotope composition of mean altered oceanic crust at ? -4.7per mil, similar to the d13C of the C degassed from the mantle at ridge axis, and discuss the global carbon budget. The total flux of C stored in the altered oceanic crust, as carbonate and organic compound, is 2.9 ± 0.4 * 10**12 molC/yr.

Geological, geochemical and biogeochemical data on boundary zones of the Atlantic Ocean

Results of studies in two biogeochemically active zones of the Atlantic Ocean (the Benguela upwelling waters and the region influenced by the Congo River run-off) are reported in the book. A multidisciplinary approach included studies of the major elements of the ocean ecosystem: sea water, plankton, suspended matter, bottom sediments, interstitial waters, aerosols, as well as a wide complex of oceanographic studies carried out under a common program. Such an approach, as well as a use of new methodical solutions led to obtaining principally new information on different aspects of oceanology.

Hydrochemical Atlas of the Arctic Ocean

Introduction: Chemical composition of water determines its physical properties and character of processes proceeding in it: freezing temperature, volume of evaporation, density, color, transparency, filtration capacity, etc. Presence of chemical elements in water solution confers waters special physical properties exerting significant influence on their circulation, creates necessary conditions for development and inhabitance of flora and fauna, and imparts to the ocean waters some chemical features that radically differ them from the land waters (Alekin & Liakhin, 1984). Hydrochemical information helps to determine elements of water circulation, convection depth, makes it easier to distinguish water masses and gives additional knowledge of climatic variability of ocean conditions. Hydrochemical information is a necessary part of biological research. Water chemical composition can be the governing characteristics determining possibility and limits of use of marine objects, both stationary and moving in sea water. Subject of investigation of hydrochemistry is study of dynamics of chemical composition, i.e. processes of its formation and hydrochemical conditions of water bodies (Alekin & Liakhin 1984). The hydrochemical processes in the Arctic Ocean are the least known. Some information on these processes can be obtained in odd publications. A generalizing study of hydrochemical conditions in the Arctic Ocean based on expeditions conducted in the years 1948-1975 has been carried out by Rusanov et al. (1979). The "Atlas of the World Ocean: the Arctic Ocean" contains a special section "Hydrochemistry" (Gorshkov, 1980). Typical vertical profiles, transects and maps for different depths - 0, 100, 300, 500, 1000, 2000, 3000 m are given in this section for the following parameters: dissolved oxygen, phosphate, silicate, pH and alkaline-chlorine coefficient. The maps were constructed using the data of expeditions conducted in the years 1948-1975. The illustrations reflect main features of distribution of the hydrochemical elements for multi-year period and represent a static image of hydrochemical conditions. Distribution of the hydrochemical elements on the ocean surface is given for two seasons - winter and summer, for the other depths are given mean annual fields. Aim of the present Atlas is description of hydrochemical conditions in the Arctic Ocean on the basis of a greater body of hydrochemical information for the years 1948-2000 and using the up-to-date methods of analysis and electronic forms of presentation of hydrochemical information. The most wide-spread characteristics determined in water samples were used as hydrochemical indices. They are: dissolved oxygen, phosphate, silicate, pH, total alkalinity, nitrite and nitrate. An important characteristics of water salt composition - "salinity" has been considered in the Oceanographic Atlas of the Arctic Ocean (1997, 1998). Presentation of the hydrochemical characteristics in this Hydrochemical Atlas is wider if compared with that of the former Atlas (Gorshkov, 1980). Maps of climatic distribution of the hydrochemical elements were constructed for all the standard depths, and seasonal variability of the hydrochemical parameters is given not only for the surface, but also for the underlying standard depths up to 400 m and including. Statistical characteristics of the hydrochemical elements are given for the first time. Detailed accuracy estimates of initial data and map construction are also given in the Atlas. Calculated values of mean-root deviations, maximum and minimum values of the parameters demonstrate limits of their variability for the analyzed period of observations. Therefore, not only investigations of chemical statics are summarized in the Atlas, but also some elements of chemical dynamics are demonstrated. Digital arrays of the hydrochemical elements obtained in nodes of a regular grid are the new form of characteristics presentation in the Atlas. It should be mentioned that the same grid and the same boxes were used in the Atlas, as those that had been used by creation of the US-Russian climatic Oceanographic Atlas. It allows to combine hydrochemical and oceanographic information of these Atlases. The first block of the digital arrays contains climatic characteristics calculated using direct observational data. These climatic characteristics were not calculated in the regions without observations, and the information arrays for these regions have gaps. The other block of climatic information in a gridded form was obtained with the help of objective analysis of observational data. Procedure of the objective analysis allowed us to obtain climatic estimates of the hydrochemical characteristics for the whole water area of the Arctic Ocean including the regions not covered by observations. Data of the objective analysis can be widely used, in particular, in hydrobiological investigations and in modeling of hydrochemical conditions of the Arctic Ocean. Array of initial measurements is a separate block. It includes all the available materials of hydrochemical observations in the form, as they were presented in different sources. While keeping in mind that this array contains some amount of perverted information, the authors of the Atlas assumed it necessary to store this information in its primary form. Methods of data quality control can be developed in future in the process of hydrochemical information accumulation. It can be supposed that attitude can vary in future to the data that were rejected according to the procedure accepted in the Atlas. The hydrochemical Atlas of the Arctic Ocean is the first specialized and electronic generalization of hydrochemical observations in the Arctic Ocean and finishes the program of joint efforts of Russian and US specialists in preparation of a number of atlases for the Arctic. The published Oceanographic Atlas (1997, 1998), Atlas of Arctic Meteorology and Climate (2000), Ice Atlas of the Arctic Ocean prepared for publication and Hydrochemical Atlas of the Arctic Ocean represent a united series of fundamental generalizations of empirical knowledge of Arctic Ocean nature at climatic level. The Hydrochemical Atlas of the Arctic Ocean was elaborated in the result of joint efforts of the SRC of the RF AARI and IARC. Dr. Ye. Nikiforov was scientific supervisor of the Atlas, Dr. R. Colony was manager on behalf of the USA and Dr. L. Timokhov - on behalf of Russia.

(Table 1) 100 m integrated temperature, chlorophyll a content and protist biomass during Leg 7 of Galathea 3 cruise (2006)

We present a study on the protozooplankton >5 µm and copepods larger than 50 µm at a series of contrasting stations across the Southern Indian Ocean (SIO). Numerically, over 80% of the copepod community across the transect was less than 650 µm in size, dominated by nauplii, and smaller copepods, while 80% of the biomass (as mg C/m**3) was larger than 1300 µm in body length. Predation by the carnivorous copepod Corycaeus sp. was estimated to be able to remove up to 2% /d of the copepods <1000 µm in size. By the help of grazing models we estimated that primary producers were mainly grazed upon by ciliates and heterotrophic dinoflagellates (40-80% /d combined) in temperate waters but appendicularians became increasingly important in the tropical waters grazing about 40% of the biomass per day. Despite their high abundance and biomass, copepods contributed less than 20% of the grazing at most stations. Secondary production was low (carbon specific egg production <0.14 /d) but typical for food limited oligotrophic oceans.

Characterization of the transposition pattern of the Ac element in Arabidopsis thaliana using endonuclease I-SceI

We have investigated physical distances and directions of transposition of the maize transposable element Ac in Arabidopsis thaliana. We prepared a transferred DNA (T-DNA) construct that carried a non-autonomous derivative of Ac with a site for cleavage by endonuclease I-SceI (designated dAc-I-RS element). Another cleavage site was also introduced into the T-DNA region outside dAc-I-RS. Three transgenic Arabidopsis plants were generated, each of which had a single copy of the T-DNA at a different chromosomal location. These transgenic plants were crossed with the Arabidopsis that carried the gene for Ac transposase and progeny in which dAc-I-RS had been transposed were isolated. After digestion of the genomic DNA of these progeny with endonuclease I-SceI, sizes of segment of DNA were determined by pulse-field gel electrophoresis. We also performed linkage analysis for the transposed elements and sites of mutations near the elements. Our results showed that 50% of all transposition events had occurred within 1,700 kb on the same chromosome, with 35% within 200 kb, and that the elements transposed in both directions on the chromosome with roughly equal probability. The data thus indicate that the Ac–Ds system is most useful for tagging of genes that are present within 200 kb of the chromosomal site of Ac in Arabidopsis. In addition, determination of the precise localization of the transposed dAc-I-RS element should definitely assist in map-based cloning of genes around insertion sites.

Effects of hydrocarbon contamination on soil microbial community and enzyme activity

Acknowledgment I would like to gratefully acknowledge the government of Saudi Arabia for the scholarship and financial support.

Insulin and insulin-like growth factor-I acutely inhibit surface translocation of growth hormone receptors in osteoblasts: A novel mechanism of growth hormone receptor regulation

We previously have demonstrated that insulin and insulin-like growth factor-I (IGF-I) down-regulate growth hormone (GH) binding in osteoblasts by reducing the number of surface GH receptors (GHRs). The present study was undertaken to investigate the mechanism of GHR down-regulation. Treatment with 5 nM insulin or IGF-I for 18 hr significantly decreased surface GH binding to 26.4 ± 2.9% and 23.0 ± 2.7% of control (mean ± SE; P < 0.05), respectively. No corresponding reductions in the mRNA level and total cellular content of GHR were found, nor was the rate of receptor internalization affected. The effects on GHR translocation were assessed by measuring the reappearance of GH binding of whole cells after trypsinization to remove the surface receptors. GH binding of control cultures significantly increased (P < 0.05) over 2 hr after trypsinization, whereas no recovery of binding activity was detected in insulin and IGF-I-treated cultures, indicating that GHR translocation was impaired. Studies on the time course of GHR down-regulation revealed that surface GH binding was reduced significantly by 3-hr treatment (P ≤ 0.0005), whereas GHR translocation was completely abolished by 75–90 min with insulin and IGF-I. The inhibition of receptor translocation by insulin, but not IGF-I, was attenuated by wortmannin. In conclusion, insulin and IGF-I down-regulated GH binding in osteoblasts by acutely impairing GHR translocation, with their effects exerted through distinct postreceptor signaling pathways.

Interaction of pigeon cytochrome c-(43–58) peptide analogs with either T cell antigen receptor or I-Ab molecule

We determined that a pigeon cytochrome c-derived peptide, p43–58, possesses two anchor residues, 46 and 54, for binding with the I-Ab molecule that are compatible to the position 1 (P1) and position 9 (P9) of the core region in the major histocompatibility complex (MHC) class II binding peptides, respectively. In the present study to analyze each binding site between P1 and P9 of p43–58 to either I-Ab or T cell antigen receptor (TCR), we investigated T cell responses to a series of peptides (P2K, P3K, P4K, P5K, P6K, P7K, and P8E) that sequentially substituted charged amino acid residues for the residues at P2 to P8 of p43–58. T cells from C57BL/10 (I-Ab) mice immunized with P4K or P6K did not mount appreciable proliferative responses to the immunogens, but those primed with other peptides (P2K, P3K, P5K, P7K, and P8E) showed substantial responses in an immunogen-specific manner. It was demonstrated by binding studies that P1 and P9 functioned as main anchors and P4 and P6 functioned as secondary anchors to I-Ab. Analyses of Vβ usage of T cell lines specific for these analogs suggested that P8 interacts with the complementarity-determining region 1 (CDR1)/CDR2 of the TCR β chain. Furthermore, sequencing of the TCR on T cell hybridomas specific for these analogs indicated that P5 interacts with the CDR3 of the TCR β chain. The present findings are consistent with the three-dimensional structure of the trimolecular complex that has been reported for TCR/peptide/MHC class I molecules.

Kinetic characterization of brush border myosin-I ATPase

Brush border myosin-I (BBM-I) is a single-headed unconventional myosin found in the microvilli of intestinal epithelial cells. We used stopped-flow kinetic analysis to measure the rate and equilibrium constants for several steps in the BBM-I ATPase cycle. We determined the rates for ATP binding to BBM-I and brush border actomyosin-I (actoBBM-I), the rate of actoBBM-I dissociation by ATP, and the rates for the steps in ADP dissociation from actoBBM-I. The rate and equilibrium constants for several of the steps in the actoBBM-I ATPase are significantly different from those of other members of the myosin superfamily. Most notably, dissociation of the actoBBM-I complex by ATP and release of ADP from actoBBM-I are both very slow. The slow rates of these steps may play a role in lengthening the time spent in force-generating states and in limiting the maximal rate of BBM-I motility. In addition, release of ADP from the actoBBM-I complex occurs in at least two steps. This study provides evidence for a member of the myosin superfamily with markedly divergent kinetic behavior.