bctsne: Calculate BC t-SNE by orthogonal gradient descent

In bcTSNE: Projected t-SNE for Batch Correction

Description

Calculate BC t-SNE by orthogonal gradient descent

Usage

bctsne(X, Z, k = 50, outDim = 2, perplexity = 30, maxIter = 1000)

Arguments

X	numeric matrix, input matrix
Z	numeric matrix, covariate matrix
k	integer of length 1, reduced dimension (number of eigenvectors)
outDim	integer of length 1, the output dimension
perplexity	numeric of length 1, the t-SNE perplexity
maxIter	integer of length 1, the maximum iterations for the BC t-SNE algorithm

Details

X should be preprocessed (e.g. PCA, centered and scaled). Z is the full model matrix, excluding the intercept.

Value

list wth the following items:

Xred

numeric matrix, the reduced dimension input to bctsne

Z

model matrix indicating batch membership

perplexity

perpelexity value used in computing t-SNE

Y

batch-corrected projection matrix

maxIter

maximum iterations used in training

Examples

## Create small simulated dataset, A, with embeded batch effects
set.seed(2731)
kRid <- 20
p    <- 100
n    <- 200

W <- matrix(rnorm(p*kRid), kRid)
S <- matrix(rnorm(n*kRid), n)
z <- sample(1:3, rep = TRUE, size = n)
Z <- model.matrix( ~ -1 + as.factor(z))
l <- matrix(rnorm(kRid*NCOL(Z)), kRid)
A <- (S - Z %*% t(l) ) %*% W

## Scale A to give input, X 
X <- scale(A)

resUnadj <- Rtsne::Rtsne(X)                 ## Standard t-SNE
resAdj   <- bctsne(X = X, Z = Z, k = 10)    ## Batch-corrected t-SNE

## Plot results, no true effects were included in the simulated data, so 
## we expect all batches to overlap with bcTSNE; batch membership indicated
## by color
plot(resUnadj$Y, col = z)
plot(resAdj$Y, col = z)

bcTSNE documentation built on Dec. 11, 2021, 9:57 a.m.