The Experts below are selected from a list of 60186 Experts worldwide ranked by ideXlab platform
Celia M T Greenwood - One of the best experts on this subject based on the ideXlab platform.
-
a sparse additive model for high dimensional interactions with an Exposure Variable
bioRxiv, 2020Co-Authors: Sahir Bhatnagar, Amanda Lovato, Celia M T Greenwood, Yi Yang, David L Olds, Michael S Kobor, Michael J Meaney, Kieran J OdonnellAbstract:Abstract A conceptual paradigm for onset of a new disease is often considered to be the result of changes in entire biological networks whose states are affected by a complex interaction of genetic and environmental factors. However, when modelling a relevant phenotype as a function of high dimensional measurements, power to estimate inter-actions is low, the number of possible interactions could be enormous and their effects may be non-linear. Existing approaches for high dimensional modelling such as the lasso might keep an interaction but remove a main effect, which is problematic for interpretation. In this work, we introduce a method called sail for detecting non-linear interactions with a key environmental or Exposure Variable in high-dimensional settings which respects either the strong or weak heredity constraints. We prove that asymptotically, our method possesses the oracle property, i.e., it performs as well as if the true model were known in advance. We develop a computationally effcient fitting algorithm with automatic tuning parameter selection, which scales to high-dimensional datasets. Through an extensive simulation study, we show that sail out-performs existing penalized regression methods in terms of prediction accuracy and support recovery when there are non-linear interactions with an Exposure Variable. We then apply sail to detect non-linear interactions between genes and a prenatal psychosocial intervention program on cognitive performance in children at 4 years of age. Results from our method show that individuals who are genetically predisposed to lower educational attainment are those who stand to benefit the most from the intervention. Our algorithms are implemented in an R package available on CRAN (https://cran.r-project.org/package=sail).
-
sparse additive interaction learning
bioRxiv, 2018Co-Authors: Sahir Bhatnagar, Amanda Lovato, Yi Yang, Celia M T GreenwoodAbstract:Abstract A conceptual paradigm for onset of a new disease is often considered to be the result of changes in entire biological networks whose states are affected by a complex interaction of genetic and environmental factors. However, when modelling a relevant phenotype as a function of high dimensional measurements, power to estimate inter-actions is low, the number of possible interactions could be enormous and their effects may be non-linear. Existing approaches for high dimensional modelling such as the lasso might keep an interaction but remove a main effect, which is problematic for interpretation. In this work, we introduce a method called sail for detecting non-linear interactions with a key environmental or Exposure Variable in high-dimensional settings which respects either the strong or weak heredity constraints. We prove that asymptotically, our method possesses the oracle property, i.e., it performs as well as if the true model were known in advance. We develop a computationally effcient fitting algorithm with automatic tuning parameter selection, which scales to high-dimensional datasets. Through an extensive simulation study, we show that sail out-performs existing penalized regression methods in terms of prediction accuracy and support recovery when there are non-linear interactions with an Exposure Variable. We then apply sail to detect non-linear interactions between genes and a prenatal psychosocial intervention program on cognitive performance in children at 4 years of age. Results from our method show that individuals who are genetically predisposed to lower educational attainment are those who stand to benefit the most from the intervention. Our algorithms are implemented in an R package available on CRAN (https://cran.r-project.org/package=sail).
Paolo Toniolo - One of the best experts on this subject based on the ideXlab platform.
-
Estimating the reliability of an Exposure Variable in the presence of confounders
Statistics in medicine, 1995Co-Authors: Mimi Y. Kim, Bernard S. Pasternack, Raymond J. Carroll, Karen L. Koenig, Paolo TonioloAbstract:In this paper we discuss estimation of the reliability of an Exposure Variable in the presence of confounders measured without error. We give an explicit formula that shows how the Exposure becomes less reliable as the degree of correlation between the Exposure and confounders increases. We also discuss biases in the corresponding logistic regression estimates and methods for correction. Data from a matched case-control study of hormone levels and risk of breast cancer are used to illustrate the methods.
Donna Spiegelman - One of the best experts on this subject based on the ideXlab platform.
-
survival analysis with measurement error in a cumulative Exposure Variable radon progeny in relation to lung cancer mortality
ISEE Conference Abstracts, 2016Co-Authors: Donna Spiegelman, Polyna Khudyakov, Jonathan M Samet, Charles L Wiggins, Xiaomei Liao, Angela L W MeisnerAbstract:Exposure Variables in occupational and environmental epidemiology are usually measured with error. This error tends to flatten the estimated Exposure-response relationship. Here, we extended the ri...
-
Power and Sample Size Calculations for Case-Control Studies of Gene-Environment Interactions with a Polytomous Exposure Variable
American journal of epidemiology, 1997Co-Authors: Ivo M. Foppa, Donna SpiegelmanAbstract:Genetic polymorphisms may appear to the epidemiologist most commonly as different levels of susceptibility to Exposure. Epidemiologic studies of heterogeneity in Exposure susceptibility aim at estimating the parameter quantifying the gene-environment interaction. In this paper, the authors use a general approach to power and sample size calculations for case-control studies, which is applicable to settings where the Exposure Variable is polytomous and where the assumption of independence between the distribution of the genotype and the environmental factor may not be met. It was found through exploration of different scenarios that in the cases explored, power calculations were relatively insensitive to assumptions about the odds ratio for the Exposure in the referent genotype category and to assumptions about the odds ratio for the genetic factor in the lowest Exposure category, yet they were relatively sensitive to assumptions about gene frequency, particularly when gene frequency was low. In general, to detect a small to moderate gene-environment interaction effect, large sample sizes are needed. Because the examples studied represent only a small subset of possible scenarios that could occur in practice, the authors encourage the use of their user-friendly Fortran program for calculating power and sample size for gene-environment interactions with Exposures grouped by quantiles that are explicitly tailored to the study at hand.
H Mallidi - One of the best experts on this subject based on the ideXlab platform.
-
steroid dosing and delirium after lung transplant surgery
Journal of Heart and Lung Transplantation, 2019Co-Authors: Gita N Mody, Keri Townsend, C Kerwin, Larios D Chavez, Steve Boukedes, A Coppolino, Steve K Singh, G Jin, David J Wolfe, H MallidiAbstract:Purpose Evaluate impact of a modified immunosuppressive regimen with decreased steroid dosing on incidence of delirium experienced by lung transplant patients. Methods We retrospectively reviewed the EMR for lung transplants between 6/14/2015 to 1/24/2017. The Exposure Variable was high (500-1000mg intraoperatively tapered to 2 mg/kg/day to 0.5mg/kg/day) versus low (500mg intraoperatively tapered to 1 mg/kg/day to 0.5mg/kg/day) steroid dosing. Covariates assessed included demographics, comorbidities, pre-transplant medications, and intraoperative Variables. Number of days with Confusion Assessment Method (CAM) positive scores and total number of high (>=3) Richmond Agitation Sedation Scale (RASS) scores in POD1-7 were recorded. We performed stepwise Variable selection and multivariate analysis using the Poisson model (for high RASS frequency) and negative binomial model (for CAM positive days). Results There were 74 lung transplants during the study period. 23 received high dose and the remainder received low dose steroids. There were no significant differences in patient characteristics aside from more patients with LAS Conclusion Delirium and agitation are common and multifactorial in lung transplant patients. A lower steroid dosing protocol appears to decrease frequency of postoperative delirium and agitation.
Michiaki Mishima - One of the best experts on this subject based on the ideXlab platform.
-
genetic relatedness of mycobacterium avium intracellulare complex isolates from patients with pulmonary mac disease and their residential soils
Clinical Microbiology and Infection, 2013Co-Authors: Kohei Fujita, Toyohiro Hirai, Koichi Maekawa, S Imai, Shuji Tatsumi, Akio Niimi, Yoshitsugu Iinuma, Satoshi Ichiyama, Michiaki MishimaAbstract:Abstract Mycobacterium avium-intracellulare complex (MAC) strains were recovered from 48.9% of residential soil samples (agricultural farms ( n = 7), residential yards ( n = 79), and planting pots ( n = 49)) of 100 pulmonary MAC patients and 35 non-infected control patients. The frequency of MAC recovery did not differ among soil types or among patients regardless of the presence of pulmonary MAC disease, infecting MAC species or period of soil Exposure. Variable numbers of tandem repeats (VNTR) analysis for MAC clinical and soil isolates revealed 78 different patterns in 47 M. avium clinical isolates and 41 soil isolates, and 53 different patterns in 18 M. intracellulare clinical isolates and 37 soil isolates. Six clinical and corresponding soil isolate pairs with an identical VNTR genotype were from case patients with high soil Exposure (≥2 h per week, 37.5% (6/16) with high Exposure compared with 0.0% (0/19) with low or no Exposure, p
