poilog_mle: Poisson-lognormal maximum likelihood estimation

In willtownes/quminorm: Quantile Normalization of Single-cell RNA-seq Read Counts Without UMIs

Description

Compute the maximum likelihood estimates for parameters of the Poisson-lognormal distribution for each column of a matrix.

Usage

poilog_mle(object, ...)

## S4 method for signature 'Matrix'
poilog_mle(object, mc.cores = 1)

## S4 method for signature 'matrix'
poilog_mle(object, mc.cores = 1)

## S4 method for signature 'SummarizedExperiment'
poilog_mle(object, assayName = "counts", mc.cores = 1)

Arguments

object	A SingleCellExperiment, SummarizedExperiment, matrix or sparse Matrix of UMI counts (non-negative integers)
...	for the generic, additional arguments to pass to object-specific methods.
mc.cores	Positive integer indicating the number of cores to use for parallel processing. See mclapply.
assayName	In case object is a SingleCellExperiment or SummarizedExperiment, the assay containing the UMI counts.

Details

This is essentially a convenience wrapper around fitpoilog with optional parallelization. WARNING: only counts from unique molecular identifiers (UMIs) should be used to fit the Poisson-lognormal, as read counts and other normalized counts are poorly fit by this distribution.

Value

a data frame whose rows correspond to the columns of the input data. The columns contain the MLEs for the parameters (mu and sig), the log-likelihood, and the BIC values. NA values indicate numerical problems with the MLE fit.

willtownes/quminorm documentation built on March 13, 2021, 2:16 a.m.