Estimate per-cell-type average expression profiles

For each gene, solves a non-negative least squares regression of bulk TPM expression across samples against per-sample cell-type fractions, giving one expression value per gene per cell type: bulk_expr[gene, sample] ~= sum_celltype fractions[sample, celltype] * profile[gene, celltype].

Usage

.estimate_expression_profiles(counts_tpm, deconv)

Arguments

counts_tpm

Gene-by-sample TPM matrix.

deconv

Sample-by-cell-type fraction matrix.

Value

A gene-by-cell-type data frame of estimated expression profiles.