pooling (version 1.1.2)

p_dfa_xerrors2: Discriminant Function Approach for Estimating Odds Ratio with Gamma Exposure Measured in Pools and Potentially Subject to Errors

Description

Archived on 7/23/18. Please use p_gdfa instead.

Usage

p_dfa_xerrors2(g, y, xtilde, c = NULL, constant_or = TRUE,
  errors = "both", integrate_tol = 1e-08,
  integrate_tol_hessian = integrate_tol, estimate_var = TRUE,
  fix_posdef = FALSE, ...)

Arguments

g

Numeric vector with pool sizes, i.e. number of members in each pool.

y

Numeric vector with poolwise Y values, coded 0 if all members are controls and 1 if all members are cases.

xtilde

Numeric vector (or list of numeric vectors, if some pools have replicates) with Xtilde values.

c

List where each element is a numeric matrix containing the C values for members of a particular pool (1 row for each member).

constant_or

Logical value for whether to assume a constant OR for X, which means that gamma_y = 0. If NULL, model is fit with and without this assumption, and likelihood ratio test is performed to test it.

errors

Character string specifying the errors that X is subject to. Choices are "neither", "processing" for processing error only, "measurement" for measurement error only, and "both".

integrate_tol

Numeric value specifying the tol input to hcubature.

integrate_tol_hessian

Same as integrate_tol, but for use when estimating the Hessian matrix only. Sometimes more precise integration (i.e. smaller tolerance) helps prevent cases where the inverse Hessian is not positive definite.

estimate_var

Logical value for whether to return variance-covariance matrix for parameter estimates.

fix_posdef

Logical value for whether to repeatedly reduce integrate_tol_hessian by factor of 5 and re-estimate Hessian to try to avoid non-positive definite variance-covariance matrix.

...

Additional arguments to pass to nlminb.

Value

List of point estimates, variance-covariance matrix, objects returned by nlminb, and AICs, for one or two models depending on constant_or. If constant_or = NULL, also returns result of a likelihood ratio test for H0: gamma_y = 0, which is equivalent to H0: log-OR is constant. If constant_or = NULL, returned objects with names ending in 1 are for model that does not assume constant log-OR, and those ending in 2 are for model that assumes constant log-OR.

References

Lyles, R.H., Van Domelen, D.R., Mitchell, E.M. and Schisterman, E.F. (2015) "A discriminant function approach to adjust for processing and measurement error When a biomarker is assayed in pooled samples." Int. J. Environ. Res. Public Health 12(11): 14723--14740.

Mitchell, E.M, Lyles, R.H., and Schisterman, E.F. (2015) "Positing, fitting, and selecting regression models for pooled biomarker data." Stat. Med 34(17): 2544--2558.

Schisterman, E.F., Vexler, A., Mumford, S.L. and Perkins, N.J. (2010) "Hybrid pooled-unpooled design for cost-efficient measurement of biomarkers." Stat. Med. 29(5): 597--613.

Whitcomb, B.W., Perkins, N.J., Zhang, Z., Ye, A., and Lyles, R. H. (2012) "Assessment of skewed exposure in case-control studies with pooling." Stat. Med. 31: 2461--2472.

Examples

# NOT RUN {
# Load dataset with (g, Y, Xtilde, C) values for 248 pools and list of C
# values for members of each pool. Xtilde values are affected by processing
# error.
data(pdat2)
dat <- pdat2$dat
c.list <- pdat2$c.list

# Estimate log-OR for X and Y adjusted for C, ignoring processing error
fit1 <- p_dfa_xerrors2(
  g = dat$g,
  y = dat$y,
  xtilde = dat$xtilde,
  c = c.list,
  errors = "neither"
)
fit1$estimates

# Repeat, but accounting for processing error.
# }
# NOT RUN {
fit2 <- p_dfa_xerrors2(
  g = dat$g,
  y = dat$y,
  xtilde = dat$xtilde,
  c = c.list,
  errors = "processing",
  control = list(trace = 1)
)
fit2$estimates
# }
# NOT RUN {
# }