est_lucid: Estimating latent clusters with multi-omics data

In USCbiostats/LUCid: Latent Unknown Clustering with Integrated Data

Description

est_lucid estimates an integrated cluster assignment of genetic effects using complete biomarker data with/without disease outcomes. Options to produce sparse solutions for cluster-specific parameter estimates under a circumstance of analyzing high-dimensional data are also provided. An IntClust object will be produced.

Usage

est_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()
)

Arguments

G	Genetic features, a matrix
CoG	Covariates to be included in the G->X path
Z	Biomarker data, a matrix, can be incomplete and have missing values
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

Value

est_lucid returns an object of list containing parameters estimates, predicted probability of latent clusters, and other features:

beta	Estimates of genetic effects, matrix
mu	Estimates of cluster-specific biomarker means, matrix
sigma	Estimates of cluster-specific biomarker covariance matrix, list
gamma	Estimates of cluster-specific disease risk, vector
pcluster	Probability of cluster, when G is null
pred	Predicted probability of belonging to each latent cluster

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.

Examples

# Integrative clustering without feature selection
set.seed(10)
IntClusFit <- est_lucid(G=G1,Z=Z1,Y=Y1,K=2,family="binary",Pred=TRUE)

## Not run: 
# Re-run the model with covariates in the G->X path
IntClusCoFit1 <- est_lucid(G=G1,CoG=CoG,Z=Z1,Y=Y1,K=2,family="binary",Pred=TRUE)

# Re-run the model with covariates in the X->Y path
IntClusCoFit2 <- est_lucid(G=G1,Z=Z1,Y=Y1,CoY=CoY,K=2,family="binary",Pred=TRUE)

# Re-run the model with covariates in both G->X and X->Y paths
IntClusCoFit3 <- est_lucid(G=G1,CoG=CoG,Z=Z1,Y=Y1,CoY=CoY,K=2,family="binary",Pred=TRUE)

# Model fit with incomplete biomarker data and covariates in both G->X & X->Y paths
IntClusCoFit3_Incomp <- est_lucid(G=G1,CoG=CoG,Z=Z1_Incomp,Y=Y1,CoY=CoY,K=2,family="binary")

## End(Not run)

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