gWQS (version 1.1.0)

gwqs: Fitting Weighted Quantile Sum regression models

Description

Fits Weighted Quantile Sum (WQS) regressions for continuous or binomial outcomes.

Usage

gwqs(formula, mix_name, data, q = 4, validation = 0.6, valid_var = NULL,
  b = 100, b1_pos = TRUE, b1_constr = FALSE, family = "gaussian",
  seed = NULL, wqs2 = FALSE, plots = FALSE, tables = FALSE)

Arguments

formula

An object of class formula specifying the relationship to be tested. If no covariates are being tested specify y ~ NULL.

mix_name

A character vector listing the variables contributing to a mixture effect.

data

The data.frame containing the variables to be included in the model.

q

An integer to specify how mixture variables will be ranked, e.g. in quartiles (q = 4), deciles (q = 10), or percentiles (q = 100). If q = NULL then the values of the mixture variables are taken (these must be standardized).

validation

Percentage of the dataset to be used to validate the model. If validation = 0 then the test dataset is used as validation dataset too.

valid_var

A character value containing the name of the variable that identifies the validation and the training dataset. You previously need to create a variable in the dataset which is equal to 1 for the observations you want to include in the validation dataset and equal to 0 for the observation you want to include in the training dataset. Assign valid_var = NULL if you want to let the function create the validation and training dataset by itself.

b

Number of bootstrap samples used in parameter estimation.

b1_pos

A logical value that determines whether weights are derived from models where the beta values were positive or negative.

b1_constr

A logial value that determines whether to apply positive (if b1_pos = TRUE) or negative (if b1_pos = FALSE) constraints in the optimization function for the weight estimation.

family

A character value, if equal to "gaussian" a linear model is implemented, if equal to "binomial" a logistic model is implemented.

seed

An integer value to fix the seed, if it is equal to NULL no seed is chosen.

wqs2

A logical value indicating whether a quadratic term should be included in the model (wqs2 = TRUE) or not (wqs2 = FALSE).

plots

A logical value indicating whether plots should be generated with the output (plots = TRUE) or not (plots = FALSE).

tables

A logical value indicating whether tables should be generated in the directory with the output (tables = TRUE) or not (tables = FALSE). A preview of the estimates of the final weights is generated in the Viewer Pane

Value

gwqs return the results of the WQS regression as well as many other objects and datasets.

fit

A glm2 object that summarizes the output of the WQS model, reflecting either a linear or logistic regression depending on how the family parameter was specified (respectively, "gaussian" or "binomial"). The summary function can be used to call and print fit data.

conv

Indicates whether the solver has converged (0) or not (1 or 2).

wb1pm

Matrix of estimated weights, mixture effect parameter estimates and the associated p-values estimated for each bootstrap iteration.

y_adj

Vector containing the y values (dependent variable) adjusted for the residuals of a fitted model adjusted for covariates.

wqs

Vector containing the wqs index for each subject.

index_b

List of vectors containing the rownames of the subjects included in each bootstrap dataset.

data_t

data.frame containing the subjects used to estimate the weights in each bootstrap.

data_v

data.frame containing the subjects used to estimate the parameters of the final model.

final_weights

data.frame containing the final weights associated to each chemical.

fit_2

It is the same as fit, but it containes the results of the regression with the wqs quadratic term. If wqs2 = FALSE, NULL is returned.

aov

Analysis of variance table to test the significance of the wqs quadratic term in the model. If wqs2 = FALSE, NULL is returned.

Details

gWQS uses the glm2 function in the glm2 package to fit the model. The glm2 package is a modified version of the glm function provided and documented in the stats package.

The solnp optimization function is used to estimate the weights in each bootstrap sample.

The seed argument specifies a fixed seed through the set.seed function.

The wqs2 argument includes a quadratic mixture effect in the linear model. In order to test the significance of this term an Analysis of Variance is executed through the anova function.

The plots argument produces two figures through the ggplot function.

References

Carrico C, Gennings C, Wheeler D, Factor-Litvak P. Characterization of a weighted quantile sum regression for highly correlated data in a risk analysis setting. J Biol Agricul Environ Stat. 2014:1-21. ISSN: 1085-7117. DOI: 10.1007/ s13253-014-0180-3. http://dx.doi.org/10.1007/s13253-014-0180-3.

Czarnota J, Gennings C, Colt JS, De Roos AJ, Cerhan JR, Severson RK, Hartge P, Ward MH, Wheeler D. 2015. Analysis of environmental chemical mixtures and non-Hodgkin lymphoma risk in the NCI-SEER NHL study. Environmental Health Perspectives, DOI:10.1289/ehp.1408630.

Czarnota J, Gennings C, Wheeler D. 2015. Assessment of weighted quantile sum regression for modeling chemical mixtures and cancer risk. Cancer Informatics, 2015:14(S2) 159-171 DOI: 10.4137/CIN.S17295.

Examples

# NOT RUN {
# we save the names of the mixture variables in the variable "toxic_chems"
toxic_chems = c("log_LBX074LA", "log_LBX099LA", "log_LBX105LA", "log_LBX118LA",
"log_LBX138LA", "log_LBX153LA", "log_LBX156LA", "log_LBX157LA", "log_LBX167LA",
"log_LBX170LA", "log_LBX180LA", "log_LBX187LA", "log_LBX189LA", "log_LBX194LA",
"log_LBX196LA", "log_LBX199LA", "log_LBXD01LA", "log_LBXD02LA", "log_LBXD03LA",
"log_LBXD04LA", "log_LBXD05LA", "log_LBXD07LA", "log_LBXF01LA", "log_LBXF02LA",
"log_LBXF03LA", "log_LBXF04LA", "log_LBXF05LA", "log_LBXF06LA", "log_LBXF07LA",
"log_LBXF08LA", "log_LBXF09LA", "log_LBXPCBLA", "log_LBXTCDLA", "log_LBXHXCLA")

# To run a linear model and save the results in the variable "results". This linear model
# (family="Gaussian") will rank/standardize variables in quartiles (q = 4), perform a
# 40/60 split of the data for training/validation (validation = 0.6), and estimate weights
# over 5 bootstrap samples (b = 3). Weights will be derived from mixture effect
# parameters that are positive (b1_pos = TRUE). A unique seed was specified (seed = 2016) so
# this model will be reproducible, and plots describing the variable weights and linear
# relationship will be generated as output (plots = TRUE). In the end tables describing the
# weights values and the model parameters with the respectively statistics are generated in
# the viewer window
results = gwqs(y ~ NULL, mix_name = toxic_chems, data = wqs_data, q = 4, validation = 0.6,
               b = 3, b1_pos = TRUE, b1_constr = FALSE, family = "gaussian", seed = 2016,
               wqs2 = FALSE, plots = TRUE, tables = TRUE)

# to test the significance of the covariates
summary(results$fit)

# }