RNAseqNet (version 0.1.5)

imputedGLMnetwork: Multiple hot-deck imputation and network inference from RNA-seq data.

Description

imputedGLMnetwork performs a multiple hot-deck imputation and infers a network for each imputed dataset with a log-linear Poisson graphical model (LLGM).

Usage

imputedGLMnetwork(X, Y, sigma, m = 50, lambdas = NULL, B = 20)

Value

S3 object of class HDpath: a list consisting of

path

a list of m data frames, each containing the adjacency matrix of the inferred network obtained from the corresonding imputed dataset. The regularization parameter is selected by StARS

efreq

a numeric matrix of size p x p, which indicates the number of times an edge has been predicted among the m inferred networks

Arguments

X

n x p numeric matrix containing RNA-seq expression with missing rows (numeric matrix or data frame)

Y

auxiliary dataset (n' x q numeric matrix or data frame)

sigma

affinity threshold for donor pool

m

number of replicates in multiple imputation (integer). Default to 50

lambdas

a sequence of decreasing positive numbers to control the regularization (numeric vector). Default to NULL

B

number of iterations for stability selection. Default to 20

Author

Alyssa Imbert, Nathalie Vialaneix,

Details

When input lambdas are null the default sequence of glmnet for the first model (the one with the first column of count as the target) is used. A common default sequence is generated for all imputed datasets using this method.

References

Imbert, A., Valsesia, A., Le Gall, C., Armenise, C., Lefebvre, G. Gourraud, P.A., Viguerie, N. and Villa-Vialaneix, N. (2018) Multiple hot-deck imputation for network inference from RNA sequencing data. Bioinformatics. tools:::Rd_expr_doi("10.1093/bioinformatics/btx819").

Examples

data(lung)
data(thyroid)
nobs <- nrow(lung)
miss_ind <- sample(1:nobs, round(0.2 * nobs), replace = FALSE)
lung[miss_ind, ] <- NA
lung <- na.omit(lung)
lambdas <- 4 * 10^(seq(0, -2, length = 10))
if (FALSE) {
lung_hdmi <- imputedGLMnetwork(lung, thyroid, sigma = 2, lambdas = lambdas,
                               m = 10, B = 5)
}