Unico: a unified model for cell-type resolution genomics from heterogeneous omics data

Abstract

Most population-scale genomic datasets collected to date consist of “bulk” samples obtained from heterogeneous tissues, reflecting mixtures of different cell types. We introduce Unico, a Unified cross-omics computational method designed to deconvolve standard two-dimensional bulk matrices (samples by features) into three-dimensional tensors (samples by features by cell types). Unico is the first principled model-based deconvolution method that is theoretically justified for any tissue-level genomic data. By deconvolving bulk gene expression and DNA methylation datasets, we demonstrate Unico’s superior performance compared to existing methods, enhancing the ability to conduct powerful, large-scale genomic studies at cell-type resolution.