SEARCH · PubMed Health
Explore indexed PubMed citations for clinical trials, systematic reviews and public health research. Read source abstracts and follow each citation to its original PubMed record.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
The distribution of interaxial angles between packed alpha-helices has been explained by a number of elegant models describing how side chains on helices can interdigitate without steric conflicts. Here I show that much of the apparent preference for particular angles is due to statistical bias and that true packing angle preferences are not well described by regular packing models.
MOTIVATION: Microarrays are capable of determining the expression levels of thousands of genes simultaneously. One important application of gene expression data is classification of samples into categories. In combination with classification methods, this technology can be useful to support clinical management decisions for individual patients, e.g. in oncology. Standard statistic methodologies in classification or prediction do not work well when the number of variables p (genes) far too exceeds the number of samples n. So, modification of existing statistical methodologies or development of new methodologies is needed for the analysis of microarray data. RESULTS: This paper proposes a new method for tumor classification using gene expression data. In this method, we first employ independent component analysis to model the gene expression data, then apply optimal scoring algorithm to classify them. Further speaking, this approach can first make full use of the high-order statistical information contained in the gene expression data. Second, this approach also employs regularized regression models to handle the situation of large numbers of correlated predictor variables. Finally, the predictive models are developed for classifying tumors based on the entire gene expression profile. To show the validity of the proposed method, we apply it to classify four DNA microarray datasets involving various human normal and tumor tissue samples. The experimental results show that the method is efficient and feasible. AVAILABILITY: Matlab scripts are available on request.
BACKGROUND: Tests to assess the change in ovarian reserve (OR) with age have been analysed only in monophasic or biphasic linear patterns. Our aim was to analyse an optimum curve that might define the relationship between different OR tests and age. METHODS: A total of 81 regularly menstruating women without a history of infertility were included in this prospective study. On cycle day 3, antral follicle (AF) counts, ovarian volume (OV), and serum FSH and estradiol levels were determined. Curve estimation was performed to determine the optimal relationship between age and OR tests. Optimum curves were also compared with monophasic and biphasic linear relationships. RESULTS: A quadratic model (y=a x x2+b x x+c) had the highest coefficient of determination for the AF count, OV and serum FSH level. The predictive power of this quadratic model was comparable with biphasic linear models for the OV and serum FSH level, but was better than that of the AF count. CONCLUSIONS: The pattern of reproductive ageing as assessed by hormonal and ultrasonographic OR tests does not appear to show an abrupt change at a certain age, but follows a continuously increasing rate of decline in the third decade of life. The changes in serum FSH levels and ultrasonographic OR tests follow a quadratic model in regularly menstruating women.
In integrated segmentation and recognition of character strings, the underlying classifier is trained to be resistant to noncharacters. We evaluate the performance of state-of-the-art pattern classifiers of this kind. First, we build a baseline numeral string recognition system with simple but effective presegmentation. The classification scores of the candidate patterns generated by presegmentation are combined to evaluate the segmentation paths and the optimal path is found using the beam search strategy. Three neural classifiers, two discriminative density models, and two support vector classifiers are evaluated. Each classifier has some variations depending on the training strategy: maximum likelihood, discriminative learning both with and without noncharacter samples. The string recognition performances are evaluated on the numeral string images of the NIST Special Database 19 and the zipcode images of the CEDAR CDROM-1. The results show that noncharacter training is crucial for neural classifiers and support vector classifiers, whereas, for the discriminative density models, the regularization of parameters is important. The string recognition results compare favorably to the best ones reported in the literature though we totally ignored the geometric context. The best results were obtained using a support vector classifier, but the neural classifiers and discriminative density models show better trade-off between accuracy and computational overhead.
After inspection of vertical sinusoidal gratings at least three distinct types of subjective or "hallucinated" patterns can be seen on a uniform test field. One type, here called horizontal streaming (H), is already well-known from the work of MacKay. A second type (V) looks like aroughly sinusoidal grating about 1-5 octaves above the adapting spitial frequency. Under optimal conditions a second vertical component appears at about 2 octaves below the adapting frequency. The third category of aftereffect consists of diagonal lines (D) at two orientations (about +/-40 degrees from vertical). The spatial-frequency band at these two orientations appears to be fairly broad, but roughly similar to the adapting frequency. The duration and strength of D increased, while V declined, at higher adapting spatial frequencies. D and V were increasing functions of adapting contrast, while H appeared abruptly only after the highest adapting contrast. H, D, and V are thus all functionally distinct. A schematic model of cortical organization is proposed to account for these phenomena. Pattern channels selective for a given orientation are grouped together with movement channels selective for the orthogonal direction. Antagonism between channels within such "modules" accounts for the streaming effect (H). Inhibition between modules tuned to different orientations and spatial frequencies accounts for the D and V effects: after adaptation of channels in one module, neighbouring module(s) are released from inhibition to produce a spurious response which is seen as a grating-like object in the adapted part of the visual field. During flickering adaptation a "halluncinated" lattice can be seen superimposed on the adapting grating. It apparently consists of Fourier components more remote from the adapting pattern than D and V are. This disinhibitory effect is strong confirmation of the inhibitory model. The regular and highly organized matrix of channels implied by these experiments may constitute a cortical hypercolumn conducting a coarse, piecewise Fourier transformation of the retinal image.
A model of nucleosome is discussed, which consists of two nucleohistone strand folds, located at the same level, similarly directed and having a rhomboid form. The folds are symmetric to each other. Four histones (H3, H2a, H2b and H4) take part in the formation of each fold. Nucleosome begins with a DNA region, bound with H1 histone and terminates with free DNA. Total sequence of histones along DNA is H1-H3-H2a-H2b-H4-H4-H2b-H2a-H3. Polypeptide chains of neighboring histones are oppositely directed and are located at opposite DNA strands. The model explains regularities of chromatin splitting under combined effect of ds-nucleases and trypsin, and of ss-nucleases. It is also in a good agreement with other properties of chromatin. Nucleosomes join to each other "side-to-side" under coincidence of terminal elements with faces of the rhomboid nucleosome structures. The model permits to explain the formation of a highest order structure--a helix of six nucleosomes, forming a fibril of 250--300 A in diameter. The degree of DNA compactness in it reaches 80--100.
BACKGROUND: Assessment of the left ventricular (LV) and the right ventricular (RV) volumes and their functions is important for prognostic prediction and clinical decision making. We compared the accuracy for quantifying the LV and the RV volumes in vitro between conventional two-dimensional echocardiography (2DE) and real-time three-dimensional echocardiography (RT3DE). METHODS: The volumes of 37 rubber-models (10 regularly shaped to simulate normal LV, 7 shaped to simulate LV with symmetric aneurysm, 8 shaped to simulate LV with asymmetric aneurysm, and 12 irregularly shaped to simulate normal RV) and 10 excised canine hearts were measured by RT3DE and 2DE. On RT3DE "full volume" imaging, the inner-surfaces of the rubber-models and canine LV and RV were outlined and the volumes were measured using 2-, 4-, 8- and 16-plane methods with the RT3DE analysis software. On 2DE imaging, the volumes were measured by the Simpson method. The LV and RV volumes measured by drained water were served as reference values, with which we compared RT3DE and 2DE data. RESULTS: In rubber models mimicking normal LV and LV with symmetric aneurysms, RT3DE results were strongly correlated with reference values (r = 0.795 - 0.998) and there was a good correlation between 2DE estimates and reference values (r = 0.715 - 0.729). There were no significant differences between RT3DE estimates, 2DE results and reference values (P > 0.05). In rubber models mimicking the RV and LV with asymmetric aneurysm, RT3DE strongly correlated with reference values (r = 0.765 - 0.988), but 2DE weakly correlated with reference values (r = 0.518 - 0.592). There were no differences between RT3DE and reference values (P > 0.05), but a significant difference between 2DE and reference values occurred (P < 0.05). For excised canine hearts, there was a strong correlation between RT3DE and reference values (r = 0.728 - 0.914), while 2DE showed a less obvious correlation (r = 0.502 - 0.615). Again, there were no significant differences between RT3DE and reference values (P > 0.05), but there was a significant difference between 2DE and reference values (P < 0.05). CONCLUSIONS: RT3DE can accurately quantify LV and RV volumes and provides a new tool to evaluate LV and RV function. For LV and RV measurements by RT3DE, 8-plane strategy is the optimum choice for accuracy and convenience.
This study examines the use of a new regularization scheme, called regional regularization, for solving the electrocardiographic inverse problem. Previous work has shown that different time frames in the cardiac cycle require varying degrees of regularization. This reflects differences in potential magnitudes, gradients, signal-to-noise ratio (SNR), and locations of electrical activity. One might expect, therefore, that a single regularization parameter and a uniform level of regularization may also be insufficient for a single potential map of a single time frame because in one map there are regions of high and low potentials and potential gradients. Regional regularization is a class of methods that subdivides a given potential map into functional "regions" based on the spatial characteristics of the potential ("spatial frequencies"). These individual regions are regularized separately and recombined into a complete map. This paper examines the hypothesis that such regionally regularized maps are more accurate than if all regions were taken together and solved with an averaged level of regularization. In a homogeneous concentric spheres model, Legendre polynomials are used to decompose a torso potential map into a set of submaps, each with a different degree of spatial variation. The original torso map is contaminated with data noise, or geometrical error or both, and regional regularization improves the epicardial potential reconstruction by up to 25% [relative error (RE)]. Regional regularization also improves the reconstructed location of peaks. A practical goal is to extend the application of this method to the realistic torso geometry, but because Legendre decomposition is limited to geometries with spherical symmetry, other methods of map decomposition must be found. Singular value decomposition (SVD) is used to decompose the maps into component parts. Its individual submaps also have different levels of spatial variation; moreover, it is generalizable to any vector, does not require spherical symmetry, and is extremely efficient numerically. Using SVD decomposition for regional regularization, significant improvement was achieved in the map quality in the presence of data noise.
Differential gene expression detection and sample classification using microarray data have received much research interest recently. Owing to the large number of genes p and small number of samples n (p >> n), microarray data analysis poses big challenges for statistical analysis. An obvious problem owing to the 'large p small n' is over-fitting. Just by chance, we are likely to find some non-differentially expressed genes that can classify the samples very well. The idea of shrinkage is to regularize the model parameters to reduce the effects of noise and produce reliable inferences. Shrinkage has been successfully applied in the microarray data analysis. The SAM statistics proposed by Tusher et al. and the 'nearest shrunken centroid' proposed by Tibshirani et al. are ad hoc shrinkage methods. Both methods are simple, intuitive and prove to be useful in empirical studies. Recently Wu proposed the penalized t/F-statistics with shrinkage by formally using the (1) penalized linear regression models for two-class microarray data, showing good performance. In this paper we systematically discussed the use of penalized regression models for analyzing microarray data. We generalize the two-class penalized t/F-statistics proposed by Wu to multi-class microarray data. We formally derive the ad hoc shrunken centroid used by Tibshirani et al. using the (1) penalized regression models. And we show that the penalized linear regression models provide a rigorous and unified statistical framework for sample classification and differential gene expression detection.
By using genome-wide markers to predict an individual's genetic potential, the introduction of Genomic Selection (GS) has transformed animal breeding. This greatly accelerated selection for complex traits by lowering reliance on drawn-out field trials, allowing for faster genetic gains in livestock. However, there is little research on the effects of genomic selection on Sahiwal cattle in India, and comparing it to the current culling or selection process is even more uncommon, particularly in nations with fewer genotyped animals. This study is an initial effort to address the aforementioned gaps in knowledge. Genomic selection was implemented in Sahiwal cattle for the 305 days milk yield using univariate animal model and the single-step Genomic Best Linear Unbiased Prediction (ssGBLUP) method. The Effective Population size (Ne) of the Sahiwal herd was calculated using genomic data and was reported for the previous generation to be 71.927. The heritability of 305 days milk yield was estimated as 0.177 ± 0.068. Genomic estimated breeding values (GEBVs) were predicted for each individual using ssGBLUP, yielding a mean prediction accuracy of 43.11%, compared with 40.88% obtained using conventional pedigree-based BLUP. Cross-validation further demonstrated superior predictive performance of ssGBLUP, with accuracies of 76.82% and 70.50% for ssGBLUP and PBLUP, respectively. To further check the effectiveness of the genomic selection methodology, we also compared the GEBVs obtained and compared it with the Expected Progeny Difference (EPD) which is being applied in our farm for culling decisions. It was seen that GEBVs obtained from ssGBLUP methodology were also in line with the conventionally used method of EPD. The use of genomic selection enables genetic studies with limited pedigree information. Additionally, the ssGBLUP methodology allows to check for pedigree errors, where family relationships are incorrectly recorded. The EPD and GEBVs were consistent with one another, indicating that genomic selection may also be utilised to support culling and selection decisions in a farm. Thus, in a conventional animal breeding program with constraint resources and an incomplete pedigree, we recommend employing the ssGBLUP model for regular genomic assessment and identification of suitable candidates to effectively carry out a genomic selection program.
OBJECTIVE: Breast cancer is the most common cancer in women, and African-American women are less likely to detect breast cancer at its early stages. Few controlled trials of interventions to increase breast self-examination (BSE) among low-income minority women have been conducted. Our objective was to test a small-group workshop intended to build BSE skills and promote BSE among low-income African-American women. METHODS: A randomized community field trial tested a BSE skills-building workshop based on social cognitive theory of behavior change compared to a matched sexually transmitted diseases prevention workshop, with one- and three-month follow-ups for assessing increased practice of BSE. RESULTS: Women who did not regularly practice BSE and participated in a BSE skills-building workshop were more likely to practice BSE than women in the comparison intervention at the one- (OR = 3.5, p = 0.04) and three-month follow-ups (OR = 4.9, p = 0.01). These results were not related to risk perceptions heightened by the intervention. Across conditions and controlling for covariates in a multivariate model, performing regular BSE was most closely associated with having received any formal BSE instruction. CONCLUSION: BSE skills building can effectively increase use of BSE among low-income African-American women who face multiple and competing health risks. The small-group experience is an important element in fostering norms for practicing BSE and enhancing BSE practices.
For many years, statistical analysis of protein databanks has led to the belief that the steric compatibility of helix interfaces may be the source of observed preferences for particular angles between neighboring helices. Several elegant models describing how side chains on helices can interdigitate without steric clashes were able to account quite reasonably for the observed distributions. However, it was later recognized that the 'bare' measured angle distribution should be corrected to avoid statistical bias.12 Disappointingly, the rescaled distributions dramatically lost their similarity with theoretical predictions, casting doubts on the validity of the geometrical assumptions and models. In this article, we elucidate a few points concerning the proper choice of a random reference distribution. In particular we demonstrate the need for corrections induced by unavoidable uncertainties in determining whether two helices are in face-to-face contact or not and their relative orientations. By using this new rescaling, we show that 'true' packing angle preferences are well described by regular packing models, thus proving that preferential angles between contacting helices do exist.
Strain rate is implemented as a stimulus for surface bone remodeling. Using idealized models for trabecular bone structures, the surface remodeling predictions using the strain rate as the stimulus are compared with the predictions using the peak strain magnitude as the stimulus. For a uniaxially loaded cruciform shape, the comparison shows that the two surface remodeling stimuli predict the same final shape under a periodic compressive load, but the two evolutionary paths to final shapes are different. Two biaxially loaded regular grid models of trabecular structure were considered, one a grid of square diamond shaped elements and the other a brick wall patterned grid. For both of these idealized trabecular structures, the comparison shows that the two surface remodeling stimuli predict the same final shape under a periodic compressive load, even from these distinctly different initial grid patterns, and the evolutionary paths to final shapes are quite different. In general the two stimuli do not predict the same remodeling and the conditions under which they do are derived. The models developed are also applied to the data from the animal experiments reported in Goldstein et al. (1991), and it is shown that the strain rate stimulus predicts bone remodeling similar to what was experimentally observed.
The transfection of mutated human insulin gene was studied using the complex of different proportion of plasmid and Liposome. Hepatic cell was used as the target cell in vitro,isolation of Hepatic cell including insulin gene was carried out by G418, the expression level of insulin in medium was measured by RIA method. The portal vein was cannulated with therapeutic gene in vivo, the blood glucose of the model was regularly examined and the insulin level was detected on the seventh day after transfection. The results showed that the target gene was transferred into the hepatic cell, expression of mature insulin was detected both in vivo and in vitro. It reached the peak 10.45microIU/ml on the 24th hour after transfection with the proportion 1:6 of plasmid and Liposome in hepatic cell. Diabetic symptom of the model was improved after transgene, the blood glucose could decrease 55% at the most.
UNLABELLED: The standard approach for evaluating FDG-PET kinetic studies is based upon an assumption that tissue within a representative region of interest (ROI) is relatively homogeneous in terms of FDG kinetics. In neoplasms and other disease states, tissue within an ROI may be grossly heterogeneous, due to adjacent infarcted tissue and other causes. We have developed a method employing two ROIs (one over the tumor and another over a "reference region") to deal with this level of heterogeneity. METHODS: The method is based on the regular FDG model but consists of six variable parameters (6P model) which uses the kinetics in the reference region to account for the normal tissue within the tumor ROI, so that the kinetic data only associated with the tumor can be estimated. Monte Carlo simulations and human PET FDG studies were used to analyze the performance of the 6P model. RESULTS: The narrower 95% confidence intervals of parameter estimates, which centered at the true tumor rate constants, and the smaller correlation matrix of the 6P model showed the better performance of the 6P model compared to the standard "homogeneous" four-parameter FDG model. Computer simulations further showed that the 6P model can accurately estimate the microparameters (rate constants: K1* (ml/min/g), k2* (min-1), k3* (min-1), k4* (min-1)) and the macroparameter (K (ml/min/g)) of tumor cells regardless of the percent weight of tumor cells in the lesions. CONCLUSIONS: The new method can produce more reliable and accurate estimates of tumor glucose metabolic rates with dynamic PET FDG studies.
Nowadays, several types of brain imaging device are available to provide images of the functional activity of the cerebral cortex based on hemodynamic, metabolic, or electromagnetic measurements. However, static images of brain regions activated during particular tasks do not convey the information of how these regions communicate with each other. In this study, advanced methods for the estimation of cortical connectivity from combined high-resolution electroencephalography (EEG) and functional magnetic resonance imaging (fMRI) data are presented. These methods include a subject's multicompartment head model (scalp, skull, dura mater, cortex) constructed from individual magnetic resonance images, multidipole source model, and regularized linear inverse source estimates of cortical current density. Determination of the priors in the resolution of the linear inverse problem was performed with the use of information from the hemodynamic responses of the cortical areas as revealed by block-designed (strength of activated voxels) fMRI. We estimate functional cortical connectivity by computing the directed transfer function (DTF) on the estimated cortical current density waveforms in regions of interest (ROIs) on the modeled cortical mantle. The proposed method was able to unveil the direction of the information flow between the cortical regions of interest, as it is directional in nature. Furthermore, this method allows to detect changes in the time course of information flow between cortical regions in different frequency bands. The reliability of these techniques was further demonstrated by elaboration of high-resolution EEG and fMRI signals collected during visually triggered finger movements in four healthy subjects. Connectivity patterns estimated for this task reveal an involvement of right parietal and bilateral premotor and prefrontal cortical areas. This cortical region involvement resembles that revealed in previous studies where visually triggered finger movements were analyzed with the use of separate EEG or fMRI measurements.
BACKGROUND: Lymphangiogenesis promotes tumor dissemination and may shape the immune contexture of cervical cancer, yet lymphangiogenesis-related prognostic stratification and its immunometabolic implications remain insufficiently defined in cervical squamous cell carcinoma and endocervical adenocarcinoma (CESC). METHODS: TCGA-CESC transcriptomes and clinical data were obtained from UCSC Xena and integrated with normal cervix tissues from the Genotype-Tissue Expression Project after batch correction. Prognostic LYMRGs were first identified from the differentially expressed set using univariable Cox proportional hazards analysis. Candidate genes were then reduced using an L1-regularized Cox model (Least Absolute Shrinkage and Selection Operator), and the remaining markers were entered into a multivariable Cox regression to obtain the final coefficients and compute an individualized risk score. The model's prognostic value was further assessed in an independent Gene Expression Omnibus dataset. In addition, expression patterns of the signature genes were leveraged for molecular subtyping of TCGA samples via non-negative matrix factorization (NMF). Immune infiltration and immunotherapy-associated characteristics were interrogated through a multi-algorithm strategy (single-sample gene set enrichment analysis, CIBERSORT, ESTIMATE, Tumor Immune Dysfunction and Exclusion (TIDE), and Immunophenoscore . Additional analyses included pathway enrichment (GSEA/GO/KEGG), drug sensitivity prediction (pRRophetic/CellMiner), and ceRNA network analysis. RESULTS: A six-gene LYMRG signature robustly stratified survival. High-risk patients had significantly worse overall survival in The Cancer Genome Atlas with AUCs of 0.819/0.801/0.801 at 1/3/5 years, and in GSE52903 (P = 0.001) with AUCs of 0.733/0.719/0.725. NMF identified two subtypes with distinct prognosis (P = 0.01) and divergent immune landscapes. Risk groups and subtypes exhibited consistent differences in immune infiltration, checkpoint expression, TIDE/IPS patterns, and pathway enrichment. Predicted chemosensitivity differed by risk group, and the ceRNA network suggested candidate upstream lncRNA regulators of the signature. CONCLUSION: A lymphangiogenesis-related six-gene model enables clinically meaningful prognostic stratification of CESC and links lymphangiogenesis programs to distinct tumor immune phenotypes and therapeutic vulnerabilities.
The use of mitomycin C during trabeculectomy is associated with a significant degree of persistent hypotony. To address this problem, we designed a scleral shield with a dry sponge glued on it for use in applying mitomycin C that reduces the penetration of the substance through the sclera. The effectiveness of the shield was tested in a rabbit model using regular sponges and sponges with the attached shield. Histopathologic examination of the eyes after 1 month revealed toxic changes in the eyes in which only a regular sponge had been used, but not in those in which the sponges with the attached scleral shield had been used. The scleral shield may help reduce the incidence of hypotony in eyes undergoing trabeculectomy with adjunctive mitomycin C.