SimData: GE, CPG, SNP, and Covariate Simulated Data

In RHclust: Vector in Partition

GE, CPG, SNP, and Covariate Simulated Data

Description

Simulated data generator containing continuous variables representing gene expression (GE) data and DNA methylation data as M-values (GPG), and categorical variable representing single nucleotide polymorphisms (SNP). GE and CPG data are simulated from a normal distribution and SNP data is simulated from a multinomial distribution. Covariate data can have uniformly distributed data or normally distributed. Cluster separation can also be distinct or non-distinct.

Usage

SimData(seed = NULL, gene = 36,
        k = c(33,33,34),
        GEbar = 5, GEsd = 0.5,
        CPGbar = 4, CPGsd = 0.5,
        SameCPG = FALSE, SameSNP = FALSE,
        ProbDist = NULL, SameGeneDist = TRUE,
        distinct=TRUE)

Arguments

seed	Set specified seed for reproducibility
gene	Numeric input that specifies the number genes
k	Cluster pattern/distribution across subjects formatted as a vector, i.e. c(33,33,34) representing 33 subjects in the first cluster, 33 in the second cluster, and 34 in the third cluster.
GEbar	Optional numeric input to change the mean distribution of GE data
GEsd	Optional numeric input to change the standard deviation of GE data
CPGbar	Optional numeric input to change the mean distribution of CPG data
CPGsd	Optional numeric input to change the standard deviation of CPG data
SameCPG	Logical value that if set to True sets the distribution of each CPG cluster around the same mean
SameSNP	Logical value that if set to True changes the probability distribution of SNPs to be the same per cluster
ProbDist	Optional list input that allows the change of SNP probability distributions per cluster. Default list stops at 5 cluster distributions. Default ProbDist = list(c(0.50,0.25,0.25), c(0.20,0.55,0.25), c(0.30,0.15,0.55), c(0.20,0.50,0.30), c(0.45,0.20,0.35))
SameGeneDist	Logical that set the covariates to follow the genetic clustering scheme where the data is uniformly distributed amongst each subgroup in each cluster. Or if set to FALSE sets the covariates to be noisy and have no discernible groups.
distinct	Logical that sets the covariate data clusters to be distinct at default, or non-distance when set to FALSE

Details

SimData creates simulated data that aims to represent real world data for gene expression (GE), DNA methylations (CPG), and single neucleotide polymorphisms (SNP). Covariate data is used to represent other potentially useful health data to further weight the genetics data. The goal of this function is to allow the user the ability to manipulate their data for testing of the VIP() or VIPcov() functions.

Value

Clusters	Vector of cluster assignment for each subject.
Vec	Numeric representation of values per cluster used for sensitivity measures.
GE	Simulated continuous data for GE. Means of each cluster changes by a factor of 5 with default standard deviation of 0.5.
CPG	Simulated continuous data for CPG. Means of each cluster changes by a factor of 4 with default standard deviation of 0.5.
SNP	Simulated categorical data for SNP.
GE_Index	Index names for GE.
CPG_Index	Index names for CPG.
SNP_Index	Index names for SNP.
Covariates	Simulated data for covariate data. By default contains 10 columns of cateogrical data and 10 columns of numeric data.

Author(s)

[e-mail]

Examples

# Generating simulated data
sd = SimData()

## Specifying seed, genes, and clusters
# sd = SimData(seed = 42, gene = 18, c(10,40,50))

# Specifying probability distribution of SNP data
l = list( c(10, 35, 55),
          c(60, 20, 20),
          c(25, 50, 25))
sd = SimData(1, g = 36, ProbDist = l)

sd = SimData(1, g = 36, c(33,33,34))

# Same SNP distribution across 2 clusters
ssnp = SimData(g = 36, k = c(10,90), SameSNP = TRUE)

# Same CpG distribution across 2 cluster
scpg = SimData(g = 36, k = c(10,90), SameCPG = TRUE)

RHclust documentation built on Aug. 15, 2023, 9:07 a.m.