Compute TPM normalized counts

Convert counts to TPM values. This function simply performs the computations for creating TPM normalized counts given a count matrix and any vector of lengths. This is useful for quickly computing scaling factors for arbitrary features and lengths. For RNA-seq, it's better to use a transcript aligner + tximport to get TPMs. Refer to https://bioconductor.org/packages/release/bioc/html/tximport.html for more information.

Usage

tpm(x, l, ...)

# Default S3 method
tpm(x, l, ...)

# S3 method for class 'matrix'
tpm(x, l)

# S3 method for class 'data.frame'
tpm(x, l)

# S3 method for class 'SummarizedExperiment'
tpm(x, l, assay = "counts")

# S3 method for class 'DelayedArray'
tpm(x, l)

Arguments

x

matrix, numeric data.frame, or SummarizedExperiment object

l

vector of feature lengths

assay

if SummarizedExperiment, what assay to use. Default = "counts"

Value

matrix containing TPM normalized counts

Examples

# Generate some fake data with fake positive feature lengths
counts <- coriell::simulate_counts()$table
gene_lengths <- rbinom(nrow(counts), 1000, 0.9)

tpms <- tpm(counts, gene_lengths)