| Type: | Package |
| Title: | Calibrated Model Fusion Approach to Combine Surrogate Markers |
| Version: | 1.0 |
| Date: | 2022-09-22 |
| Description: | Uses a calibrated model fusion approach to optimally combine multiple surrogate markers. Specifically, two initial estimates of optimal composite scores of the markers are obtained; the optimal calibrated combination of the two estimated scores is then constructed which ensures both validity of the final combined score and optimality with respect to the proportion of treatment effect explained (PTE) by the final combined score. The primary function, pte.estimate.multiple(), estimates the PTE of the identified combination of multiple surrogate markers. Details are described in Wang et al (2022) <doi:10.1111/biom.13677>. |
| License: | GPL-2 | GPL-3 [expanded from: GPL] |
| Imports: | splines, MASS, stats |
| NeedsCompilation: | no |
| Packaged: | 2022-09-23 01:33:03 UTC; parastlm |
| Author: | Xuan Wang [aut], Layla Parast [cre] |
| Maintainer: | Layla Parast <parast@austin.utexas.edu> |
| Depends: | R (≥ 3.5.0) |
| Repository: | CRAN |
| Date/Publication: | 2022-09-23 15:20:05 UTC |
Example data
Description
Example data
Usage
data("example.data")Format
A list with 3 elements:
sobthe surrogate markers
yobthe primary outcome
aobthe treatment indicator
Examples
data(example.data)
names(example.data)
Generate bootstrap sample
Description
Generate bootstrap sample
Usage
gen.bootstrap.weights(n, num.perturb = 500)
Arguments
n |
sample size |
num.perturb |
number of replicates/resamples |
Value
matrix with n rows and num.perturb columns of indeces
Estimates the proportion of treatment effect explained by the optimal combination of multiple surrogate markers using a calibrated model fusion approach
Description
Estimates the proportion of treatment effect explained by the optimal combination of multiple surrogate markers using a calibrated model fusion approach
Usage
pte.estimate.multiple(sob, yob, aob, var = TRUE, rep = 500)
Arguments
sob |
surrogates |
yob |
primary outcome, y |
aob |
treatment indicator |
var |
TRUE or FALSE, if variance/SE of PTE is being requested |
rep |
if var is TRUE, number of resampled draws to use for bootstrap |
Value
pte.es |
Estimate of the proportion of treatment effect explained (PTE) |
pte.se |
if var = TRUE, estimate of the standard error of the PTE |
References
Wang, X., Parast, L., Han, L., Tian, L., & Cai, T. (2022). Robust approach to combining multiple markers to improve surrogacy. Biometrics, In press.
Examples
data(example.data)
out=pte.estimate.multiple(sob=example.data$sob, yob=example.data$yob,
aob=example.data$aob, var = FALSE)
out
Estimates quantities using resampled data
Description
Estimates quantities using resampled data
Usage
resam(index, yob, sob, aob, n)
Arguments
index |
index |
yob |
y |
sob |
surrogates |
aob |
treatment |
n |
n |
Value
Outputs parametric estimate, additive linear estimate, and convex combination estimate