PRIME: Probabilistic imputation to reduce dropout effects in single...

In jeonglab/prime:

Description

Probabilistic imputation to reduce dropout effects in single cell sequencing data

Usage

PRIME(sc_cnt, max_it = 5, alp = 1, err_max = 0.05)

Arguments

sc_cnt	(M x N) dimensional input matrix. This matrix sould be in count scale (not a log scale) M: number of genes and N: Number of cells
max_it	Maximum number of iteration
alp	Shape parameter for a sigmoid function (probabilistic weight)
err_max	Error tolerance to stop the iteration

Value

Normalized imputation result in count scale

Author(s)

Hyundoo Jeong

References

Hyundoo Jeong and Zhandong Liu

PRIME: a probabilistic imputation method to reduce dropout effects in single cell RNA sequencing

Examples

data(testdata)
PRIME_res <- PRIME(testdata)
pca_res <-prcomp(t(log10(1+PRIME_res)))
plot(pca_res$x[,1], pca_res$x[,2], bg = c("red", "blue","green")[factor(label)],  type = "p", pch = 21, xlab = "PC1", ylab = "PC2")
legend("topleft", title="Cell types", c("G1","S","G2M"), fill = c("red", "blue","green"), horiz = TRUE)

jeonglab/prime documentation built on May 7, 2019, 6:58 p.m.