Application of Logistic Regression and Logistic Discrimination in Medical Decision Making

The logit linear model has been shown to be very useful for investigations in many fields of biology and medicine. One kind of application of this model is that of logistic discriminant analysis. To conquer some theoretical disadvantages of logistic discriminant analysis and Fisher's linear discriminant analysis too, it has been recommended to augment their model equations with appropriate interaction terms. The present paper discusses experiences with epidemiologic data going this way.     

Multivariate analysis of bacterial volatile compound profiles for discrimination between selected species and strains in vitro

Volatile compounds (VCs) are produced by all microorganisms as part of their normal metabolism. The aim of this study was to determine whether bacterial VC profiles could be used to discriminate between selected bacterial species and strains in vitro.Selected Ion Flow Tube Mass Spectrometry (SIFT-MS) was used to quantify the concentration of 23 microbial VCs within the head-space of various bacterial monocultures, during both the logarithmic and stationary growth phases. In comparison with existing techniques, SIFT-MS enables quantitative, high throughput, real-time head-space analysis to be performed, without need for sample preparation. The results show that most VCs were produced by > 1 bacterial species or strain, and some were produced by all strains tested. Multivariate analysis using similarity matrices, cluster analysis and multidimensional scaling (MDS) was used to determine whether there was a characteristic VC profile at either the species or strain level. Significant discrimination of all bacterial species and strains was achieved by analysing the VC profiles, and the relative similarity of VC profiles could be differentiated in 2 or 3 dimensional space. This study has shown that there are significant differences in the volatile profiles obtained from various bacterial monocultures grown in vitro, and that the analysis techniques herein employed have the potential to differentiate samples at the strain level.     

Fisher判别分析法在川金丝猴亚种鉴定中的应用

根据川金丝猴头骨标本的性征指标数据,应用Fisher判别分析方法,对川金丝猴的亚种进行了判别分类．     

Analysis of the disintegrating group and gradient structure in Swiss riparian forests

The differentiation of river flood plain vegetation is predominantly determined by the inundation frequency and the type and age of the substrate. A comparison of data published in 1958 with recent relevés from all parts of Switzerland indicate that the hitherto known gradient- and group structure is disintegrating into multiple floristically isolated stands as a result of waterflow management constructions. In view of further vegetation mapping, a two-phase classification is proposed where the core of the gradients and groups is isolated and analysed first. The outliers are then re-attached to serve as possible indicators of the existence of variation not sufficiently described by the data. To structure the tedious analytical calculations involved, a hierarchical procedure has been developed.Abbreviations AOC = Analysis of concentration - CA = Correspondence analysis - PCA = Principal components analysis     

巢式设计的多元分析

本文把经典巢式设计的统计分析方法由单变量推广到多变量,使得试验或调查的多性状数据能综合在一起进行分析,以得出考虑了性状间相关性在内的统计结论并用一烟草栽培试验数据示明具体步骤。     

New Procedure for Selection of Variables in Location Model for Mixed Variable Discrimination

A new method for the choice of variables with the greatest discriminatory power in the location model for mixed variable discriminant analysis is presented in the paper. The procedure based on the multivariate discriminatory measure enables a simultaneous reduction of the number of discrete and continuous variables. The introduced criterion can be used for both optimal or step-wise selection of variable subset. As an example the results of the stepwise variable selection for some medical data are presented in the paper.     

Bacterioplankton community structure in a eutrophic lake in relation to water chemistry

Bacteria are thought to be closely involved in the biogeochemical cycling of the nutrient elements in freshwater ecosystems. In contrast, little is known about the relationship between the bacterial community structure and the environmental factors. In order to investigate the spatial variety of bacteriplankton community structure in Lake Xuanwu (Nanjing, China) and the effects of water quality on the microbial community composition, denaturing gradient gel electrophoresis (DGGE) and multivariate statistical analysis were carried out. Eight major factors of water quality such as total organic carbon, ammonium, nitrate, nitrite, total nitrogen, total phosphorus, dissolved oxygen and pH were measured in eight different samples. The significant differences of water characteristics among different sites were revealed by principal component analysis. Results of multidimensional scaling analysis demonstrated that lake water quality played a crucial role in bacterioplankton community composition. Canonical correspondence analysis was carried out to infer the relationship between environmental variables and bacterial community structure. An influence of total nitrogen, ammonium and pH on the bacterial community was observed. The sequencing analysis results of excised DGGE bands revealed that Bacteroidetes and Proteobacteria were the dominant bacterial groups in Lake Xuanwu.     

Multivariate analysis of anthropometric data and classifications of British Columbian natives

Anthropometric data collected in native populations of British Columbia in the late 19th century by Franz Boas were analyzed by two multivariate techniques. Multivariate analysis of variance was used to test physical classificatory units devised by Boas and an ad hoc classification based on local cultural units. Both were found to have some empirical validity. Mahalanobis' D (Mahalanobis, '30) was computed between pairs of local groups, for both sexes. From these a matrix of differences was prepared and diagrams drawn to illustrate phenetic relationships among samples. By this means one cluster of groups, Interior B.C. peoples, could be distinguished and other local samples appeared distinctly different from each other. It was concluded that in the absence of genealogical data by which to identify local populations, local cultural units are preferable to more inclusive units for making empirical comparisons and classifications.     

Discrimination of nestmate workers and drones in honeybees

Summary Discrimination of nestmates from non-nestmates has mainly been investigated in female social insects. Little is known about discrimination of males. Here we show that under natural conditions at the nest entrance, honeybee workers can discriminate nestmate drones from non-nestmate drones as effectively as they can discriminate nestmate workers from non-nestmate workers.     

Vertical distribution of bacterial community structure in the sediments of two eutrophic lakes revealed by denaturing gradient gel electrophoresis (DGGE) and multivariate analysis techniques

Vertical distribution of bacterial community structure was investigated in the sediments of two eutrophic lakes of China, Lake Taihu and Lake Xuanwu. Profiles of bacterial communities were generated using a molecular fingerprinting technique, denaturing gradient gel electrophoresis (DGGE) followed by DNA sequence analysis, and the results were interpreted with multivariate statistical analysis. To assess changes in the genetic diversity of bacterial communities with changing depth, DGGE banding patterns were analysed by cluster analysis. Distinct clusters were recognized in different sampling stations of Lake Taihu. Canonical correspondence analysis (CCA) was carried out to infer the relationship between environmental variables and bacterial community structure. DGGE samples collected at the same sampling site clustered together in both lakes. Total phosphorus, organic matter and pH were considered to be the key factors driving the changes in bacterial community composition.     

Cladistic and phenetic studies of genera of Afrotropical Cryptobiina (Coleoptera,Staphylinidae, Paederinae) II. The phenetic approach

The phenetics of 12 genera of Afrotropical Cryptobiina have been analysed by multivariate statistics. Three analyses (principal component analysis, canonical discriminant analysis and cluster analysis) have been used to determine the relative positions of the taxa in a morphospace. These analyses have allowed us to confirm the generic divisions based on alpha-systematic and phylogenetic grounds. While phenetic systematics and cladistic systematics have often been opposed, this study shows that both approaches are, in the end, complementary. Phenetical analysis allow, in addition to a quantitative definition of the genera, the isolation of ethological or ecological characters, while the cladistic approch gives to them precision and a phylogenetic sense. The three types of analyses point out the importance in Cryptobiina of the size, length of the scapus and width of the neck, all three characters which were also designated as significant characters by the phylogenetic analysis.     

DNA diversity of human populations from Eastern Europe and Siberia studied by multilocus DNA fingerprinting

We used DNA fingerprinting with M13 phage DNA as a probe to estimate the degree of genomic variability and genetic relationships in a heterogeneous group of 13 populations from Eastern Europe and Siberia. The popultaions belong to three language families: Indo-European (Slavonic: Russians, Byelorussians), Uralic (Finno-Ugric: Maris, Mordvinians, Udmurts), and Altaic (Turkic: Bashkirs, Tatars, Chuvashes, Yakuts). Multivariate statistical analyses were used (multidimensional scaling, cluster, and multiple correspondence analyses), and coefficients of gene differentiation (Gst) were evaluated. The level of interpopulation subdivision in the various ethnic groups appeared to be different: the Byelorussian populations revealed no regional differences, in contrast to the Bashkir populations, which formed a heterogeneous group. The populations subdivided into three general clusters: Slavonic populations formed a separate tight cluster characterized by a minimal level of interpopulation diversity, Bashkir and Yakut populations formed the second cluster, and the Finno-Ugric and several populations of the Turkic linguistic groups formed the third cluster. The robustness of these results obtained by different statistical data treatments reveals that multilocus DNA fingerprinting can be reliably used for population studies.Communicated by G. P. Georgiev     

应用一元线性模型理论确定立地条件主导因子——朝阳低山丘陵区农田防护林立地条件类型的划分

本文用一元线性模型理论分析了辽宁西部朝阳、建平、喀左和北票(市)4县17个乡范围内的农田防护林主要树种——杨树(Populus spp.)、白榆(Ulumus pumila)的优势高与地形地貌、土壤肥力等立地条件因子的关系,确定地貌类型、坡位和土层厚度是本区农田防护林立地条件的主导因子,然后以地貌类型为单元将本区农田防护林的立地条件划分为3个大类型。在此基础上,再用坡位和土层厚度分级与3个大类型组合细分为6个基本类型。最后确定各立地条件类型的适宜树种,并定量地评价主要树种对各立地条件类型的适宜性。     

Testing for Odor Discrimination and Habituation in Mice

This video demonstrates a technique to establish the presence of a normally functioning olfactory system in a mouse. The test helps determine whether the mouse can discriminate between non-social odors and social odors, whether the mouse habituates to a repeatedly presented odor, and whether the mouse demonstrates dishabituation when presented with a novel odor. Since many social behavior tests measure the experimental animal’s response to a familiar or novel mouse, false positives can be avoided by establishing that the animals can detect and discriminate between social odors. There are similar considerations in learning tests such as fear conditioning that use odor to create a novel environment or olfactory cues as an associative stimulus. Deficits in the olfactory system would impair the ability to distinguish between contexts and to form an association with an olfactory cue during fear conditioning. In the odor habitation/dishabituation test, the mouse is repeatedly presented with several odors. Each odor is presented three times for two minutes. The investigator records the sniffing time directed towards the odor as the measurement of olfactory responsiveness. A typical mouse shows a decrease in response to the odor over repeated presentations (habituation). The experimenter then presents a novel odor that elicits increased sniffing towards the new odor (dishabituation). After repeated presentation of the novel odor the animal again shows habituation. This protocol involves the presentation of water, two or more non-social odors, and two social odors. In addition to reducing experimental confounds, this test can provide information on the function of the olfactory systems of new knockout, knock-in, and conditional knockout mouse lines.     

Discrimination of random dot texture patterns in bluegill sunfish,Lepomis macrochirus

In order to study the ability of fish to perceive and distinguish textures visually, bluegill sunfish (Lepomis macrochirus) were trained to discriminate between pairs of artificial texture patterns. Random dot patterns with different statistical dot distributions were presented to the fish as artificial texture patterns. The results indicate that bluegills have the ability to discriminate many pairs of patterns with different statistical features of dot distributions which have different appearance in texture. This suggests that texture could be one of the important visual features bluegill sunfish recognize and utilize.     

Normal Approximations of Exact Tests in Configural Frequency Analysis

The most common tests for types and antitypes in configural frequency analysis are normal approximations of exact tests. In the paper such statistics under the complete independence model and under the fixed margins model are discussed. It turns out that these test statistics are not acceptable when the number of simultaneously performed tests is large or when the expected frequencies are small. In these cases, the use of exact tests is advocated and some existing computer programs for such tests are indicated. A normal approximation based on the strong version of the De Moivre-Laplace limit theorem is also discussed. Empirical examples are given from longitudinal data describing psychological development of boys.     

Discrimination of oilseed rape volatiles by the honeybee: combined chemical and biological approaches

Honeybees (Apis mellifera L.) were individually subjected to a classical conditioning procedure in order to obtain an olfactory conditioned proboscis extension response. To relate the behavioural response directly to antennal detection abilities, a technique was developped for coupling proboscis extension responses and electroantennogram recordings, with the stimulation being provided by the effluent of a gas chromatograph (GC). Bees were conditioned with a six-component mixture being part of oilseed rape (Brassica napus L.) floral volatiles, and tested with the individual components separated by GC. Responses of the conditioned bees were compared to those of unconditioned bees. No behavioural response was obtained in the control group, neither to the individual components nor to the mixture. Conditioning induced behavioural responses for three components, and an increase of electroantennogram responses for all components. A second experiment was conducted with an air entrainment extract of oilseed rape flower volatiles. Behavioural responses of conditioned and unconditioned bees were recorded. Responses obtained from conditioned bees tested with the air entrainment extract showed six groups of behaviourally active GC peaks. Unconditioned bees showed the same pattern of responses but at a lower level. The coupled technique described here appears to be a reliable tool for locating active components in a synthetic as well as in a natural mixture of floral volatiles. The effects of conditioning on odour discrimination and on its sensory correlates are discussed.