boot_lucid: Bootstrap method to estimate variability of latent clusters

In USCbiostats/LUCid: Latent Unknown Clustering with Integrated Data

Description

boot_lucid provides SEs of parameter estimates from a LUCID model through bootstrapping.

Usage

boot_lucid(
  G = NULL,
  CoG = NULL,
  Z = NULL,
  Y,
  CoY = NULL,
  useY = TRUE,
  family = "binary",
  K = 2,
  Pred = TRUE,
  initial = def_initial(),
  itr_tol = def_tol(),
  tunepar = def_tune(),
  R = 100,
  DeltaE = TRUE,
  NCPUs = detectCores() - 1
)

Arguments

G	Genetic features, a matrix
CoG	Covariates to be included in the G->X path
Z	Biomarker data, a matrix
Y	Disease outcome, a vector
CoY	Covariates to be included in the X->Y path
useY	Using Y or not, default is TRUE
family	"binary" or "normal" for Y
K	Pre-specified # of latent clusters, default is 2
Pred	Flag to compute posterior probability of latent cluster with fitted model, default is TRUE
initial	A list of initial model parameters will be returned for integrative clustering
itr_tol	A list of tolerance settings will be returned for integrative clustering
tunepar	A list of tuning parameters and settings will be returned for integrative clustering
R	The number of bootstrap replicates, default is 100
DeltaE	Flag to return the difference in parameter estimate across latent clusters, default is TRUE
NCPUs	The number of processes to be used in parallel computing, default is total number of cores minus 1

Value

boot_lucid returns an object of list containing a "boot" class object of LUCID fit and a summary of bootstrap results:

Bootstrap	an object of "boot" class after bootstrapping a LUCID model
Results	A summary of bootstrap includes original estimate, a bias of the bootstrap estimate, standard error of the bootstrap estimate, and three types of bootstrap confidence intervals based on normal approximation, basic, percentile bootstrap methods

Author(s)

Cheng Peng, Zhao Yang, David V. Conti

References

Cheng Peng, Jun Wang, Isaac Asante, Stan Louie, Ran Jin, Lida Chatzi, Graham Casey, Duncan C Thomas, David V Conti, A Latent Unknown Clustering Integrating Multi-Omics Data (LUCID) with Phenotypic Traits, Bioinformatics, , btz667, https://doi.org/10.1093/bioinformatics/btz667.

Efron, B., Tibshirani, R. Bootstrap Methods for Standard Errors, Confidence Intervals, and Other Measures of Statistical Accuracy. Statist. Sci. 1 (1986), no. 1, 54–75. doi:10.1214/ss/1177013815.

Examples

## Not run: 
boot_lucid(G = G1, CoG = CoG, Z = Z1, Y = Y1, CoY = CoY, family = "binary", R=500)

## End(Not run)

USCbiostats/LUCid documentation built on Feb. 22, 2020, 8:57 p.m.