scholarly journals Generalized linear modeling with regularization for detecting common disease rare haplotype association

2009 ◽  
Vol 33 (4) ◽  
pp. 308-316 ◽  
Author(s):  
Wei Guo ◽  
Shili Lin
Keyword(s):  
Common Disease ◽  
Linear Modeling ◽  
Rare Haplotype ◽  
Haplotype Association ◽  
Generalized Linear Modeling

City Children’s Nature Knowledge and Contact: It Is Not Just About Biodiversity Provision

2017 ◽  
Vol 50 (10) ◽  
pp. 1145-1171 ◽  
Author(s):  
Claire Freeman ◽  
Aviva Stein ◽  
Kathryn Hand ◽  
Yolanda van Heezik
Keyword(s):  
New Zealand ◽  
Environmental Factors ◽  
Ethnic Group ◽  
Social Factors ◽  
Green Spaces ◽  
Social Connections ◽  
Relative Importance ◽  
Linear Modeling ◽  
Generalized Linear Modeling ◽  
Nature Contact

Much attention has been directed at the perceived decline in city children’s contact with nature. We used a child-centric approach to assess neighborhood nature knowledge in 187 children aged 9 to 11 years, from different socioeconomic and ethnic groups in three New Zealand cities. We evaluated the relative importance of social (independence, gender, social connections, deprivation, age) and environmental factors (biodiversity) in explaining variation in knowledge at a scale relevant to each child’s independent movements. Our biodiversity evaluation reflected the natural dimensions of the habitats where children interacted with nature. Generalized linear modeling identified ethnicity as having the strongest association with nature knowledge. Within each ethnic group, social factors were most important (independence, social connections, deprivation) except for Pākehā/NZ European children, where local biodiversity was most important. Enhancing biodiversity values of private green spaces (yards) would be effective in facilitating opportunities to experience nature, which is fundamental to supporting nature contact.

scholarly journals An Improved Version of Logistic Bayesian LASSO for Detecting Rare Haplotype-Environment Interactions with Application to Lung Cancer

2015 ◽  
Vol 14s2 ◽  
pp. CIN.S17290 ◽  
Author(s):  
Yuan Zhang ◽  
Swati Biswas
Keyword(s):  
Lung Cancer ◽  
Rare Variants ◽  
Computation Time ◽  
Rare Haplotype ◽  
Cancer Dataset ◽  
Bayesian Lasso ◽  
Haplotype Association ◽  
Gene Environment ◽  
Multiple Covariates ◽  
Logistic Bayesian Lasso

The importance of haplotype association and gene-environment interactions (GxE) in the context of rare variants has been underlined in voluminous literature. Recently, a software based on logistic Bayesian LASSO (LBL) was proposed for detecting GxE, where G is a rare (or common) haplotype variant (rHTV)-it is called LBL-GxE. However, it required relatively long computation time and could handle only one environmental covariate with two levels. Here we propose an improved version of LBL-GxE, which is not only computationally faster but can also handle multiple covariates, each with multiple levels. We also discuss details of the software, including input, output, and some options. We apply LBL-GxE to a lung cancer dataset and find a rare haplotype with protective effect for current smokers. Our results indicate that LBL-GxE, especially with the improvements proposed here, is a useful and computationally viable tool for investigating rare haplotype interactions.

Simple estimators of salmonid escapement and its variance using a new area-under-the-curve method

2012 ◽  
Vol 69 (6) ◽  
pp. 1002-1015 ◽  
Author(s):  
Russell B. Millar ◽  
Sam McKechnie ◽  
Chris E. Jordan
Keyword(s):  
Direct Method ◽  
Area Under The Curve ◽  
Variance Estimator ◽  
Linear Modeling ◽  
Generalized Linear Modeling ◽  
Modeling Software ◽  
Curve Method ◽  
Ml Estimator ◽  
Alternative Estimator ◽  
A Direct Method

If observed numbers of spawning salmon are plotted against sampling date, then the area under the curve (AUC) gives an estimate of spawner-days. Dividing AUC by spawner lifetime and adjusting for observer efficiency gives an estimate of spawner escapement. In particular, the trapezoidal form of AUC estimator has been widely used over the last two decades, despite the absence of a direct method for calculating its variance. For this reason, an alternative estimator of escapement was developed using a maximum likelihood (ML) model of spawner arrivals. However, this alternative has not been widely used, perhaps because of its complexity and concerns over validity of assumptions. Here, a simpler ML approach is used to estimate AUC by fitting a model directly to spawner numbers. It can be fitted using existing generalized linear modeling software and provides an explicit variance estimator for AUC and escapement. Simulations show that it has consistently higher coverage than the existing ML estimator. However, the trapezoidal AUC estimator was generally a slightly more precise estimator of escapement than either of the ML approaches.

scholarly journals Quality of Information Regarding Repair Restorations on Dentist Websites: Systematic Search and Analysis (Preprint)

2019 ◽  
Author(s):  
Philipp Kanzow ◽  
Amelie Friederike Büttcher ◽  
Annette Wiegand ◽  
Falk Schwendicke
Keyword(s):  
Decision Making ◽  
Risk Of Bias ◽  
Shared Decision ◽  
Linear Modeling ◽  
Quality Of Information ◽  
Generalized Linear Modeling ◽  
Functional Aspects ◽  
German Speaking ◽  
Generic Quality

BACKGROUND Repairing instead of replacing partially defective dental restorations represents a minimally invasive treatment concept, and repairs are associated with advantages over complete restoration replacement. To participate in the shared decision-making process when facing partially defective restorations, patients need to be aware of the indications, limitations, and advantages or disadvantages of repairs. Patients are increasingly using the internet to gain health information like this online. OBJECTIVE We aimed to assess the quality of German-speaking dentist websites on repairs of partially defective restorations. METHODS Three electronic search engines were used to identify German-speaking websites of dental practices mentioning repairs. Regarding information on repairs, websites were assessed for (1) technical and functional aspects, (2) comprehensiveness of information, and (3) generic quality and risk of bias. Domains 1 and 3 were scored using validated tools (LIDA and DISCERN). Comprehensiveness was assessed using a criterion checklist related to evidence, advantages and disadvantages, restorations and defects suitable for repairs, and information regarding technical implementation. Generalized linear modeling was used to assess the impact of practice-specific parameters (practice location, practice setting, dental society membership, and year of examination or license to practice dentistry) on the quality of information. An overall quality score was calculated by averaging the quality scores of all three domains and used as primary outcome parameter. Quality scores of all three domains were also assessed individually and used as secondary outcomes. RESULTS Fifty websites were included. The median score of quality of information was 23.2% (interquartile range [IQR] 21.7%-26.2%). Technical and functional aspects (55.2% [IQR 51.7%-58.6%]) showed significantly higher quality than comprehensiveness of information (8.3% [IQR 8.3%-16.7%]) and generic quality and risk of bias (3.6% [IQR 0.0%-7.1%]; <i>P</i>&lt;.001/Wilcoxon). Quality scores were not related to practice-specific parameters (<i>P</i>&gt;.05/generalized linear modeling). CONCLUSIONS The quality of German-speaking dentist websites on repairs was limited. Despite sufficient technical and functional quality, the provided information was neither comprehensive nor trustworthy. There is great need to improve the quality of information to fully and reliably inform patients, thereby allowing shared decision making.

scholarly journals Dystonia & tremor: A cross-sectional study of the dystonia coalition cohort

Neurology ◽  
2020 ◽  
pp. 10.1212/WNL.0000000000011049
Author(s):  
Aasef G. Shaikh ◽  
Sinem Balta Beylergil ◽  
Laura Scorr ◽  
Gamze Kilic-Berkmen ◽  
Alan Freeman ◽  
...  
Keyword(s):  
Clinical Manifestations ◽  
Cross Sectional Study ◽  
Body Region ◽  
Linear Modeling ◽  
Cross Sectional ◽  
Factors Associated ◽  
Generalized Linear Modeling ◽  
Apparent Prevalence ◽  
Different Types ◽  
The Many

Objective:To assess the clinical manifestations and predictors of different types of tremors in a individuals with different types of isolated dystonia.Methods:Clinical manifestations of tremor were assessed in a multicenter, international cross-sectional, cohort study of 2362 individuals with all types of isolated dystonia (focal, segmental, multifocal and generalized) recruited through the Dystonia Coalition.Results:Methodical and standardized assessments of all subjects in this cohort revealed the overall prevalence of any type of tremor was 53.3%. The prevalence of dystonic tremor varied from 36.9-48.4%, depending on criteria used to define it. To identify the factors associated with tremors in dystonia, the data were analyzed by generalized linear modeling and cluster analyses. Generalized linear modeling indicated two of the strongest factors associated with tremor included body region affected by dystonia and recruitment center. Tremor was also associated with severity of dystonia and duration of dystonia, but not with sex or race. The cluster analysis distinguished eight subgroups within the whole cohort; defined largely by body region affected with dystonia, and secondarily by other clinical characteristics.Conclusion:The large number of cases evaluated by an international team of movement disorder experts facilitated the dissection of several important factors that influence the apparent prevalence and phenomenology of tremor in dystonia. These results are valuable for understanding the many differences reported in prior studies, and for guiding future studies of the nosology of tremor and dystonia.

Sign in / Sign up

Export Citation Format

Share Document