Convert counts to transcripts per million — normalize

Gene lengths may be supplied directly, read from one rowData column, or read from an aligned assay such as the sample-specific effective-length assay produced by import_tximport(). No external preprocessing step is required.

Usage

normalize_tpm(x, experiment, lengths = NULL, assay = "counts")

Arguments

x: A MultiAssayExperiment object.
experiment: Optional experiment name.
lengths: A positive numeric vector, a feature-named numeric vector, a feature-by-sample matrix, or the name of a rowData column or assay. When NULL, an assay named length (as produced by import_tximport()) is used.
assay: Assay name or one-based assay index.

Value

A finite feature-by-sample TPM matrix whose columns sum to one million.