AdRoit is an accurate and robust method to infer complex transcriptome composition

Tao Yang,Christina Adler,Yu Bai,Michael Schaner,Gurinder S Atwal,Jinrang Kim,Wen Fury,Michael Lacroix-Fralish,Robert Breese,Lynn E Macdonald,Nicole Alessandri-Haber

doi:10.1038/s42003-021-02739-1

Abstract

Bulk RNA sequencing provides the opportunity to understand biology at the whole transcriptome level without the prohibitive cost of single cell profiling. Advances in spatial transcriptomics enable to dissect tissue organization and function by genome-wide gene expressions. However, the readout of both technologies is the overall gene expression across potentially many cell types without directly providing the information of cell type constitution. Although several in-silico approaches have been proposed to deconvolute RNA-Seq data composed of multiple cell types, many suffer a deterioration of performance in complex tissues. Here we present AdRoit, an accurate and robust method to infer the cell composition from transcriptome data of mixed cell types. AdRoit uses gene expression profiles obtained from single cell RNA sequencing as a reference. It employs an adaptive learning approach to alleviate the sequencing technique difference between the single cell and the bulk (or spatial) transcriptome data, enhancing cross-platform readout comparability. Our systematic benchmarking and applications, which include deconvoluting complex mixtures that encompass 30 cell types, demonstrate its preferable sensitivity and specificity compared to many existing methods as well as its utilities. In addition, AdRoit is computationally efficient and runs orders of magnitude faster than most methods.

Highlights

Bulk RNA sequencing provides the opportunity to understand biology at the whole transcriptome level without the prohibitive cost of single cell profiling
The most common strategy is to conduct a statistical inference through the maximum likelihood estimation (MLE)[25] or the maximum a posterior probability estimation (MAP)[26] on a constrained linear regression framework, wherein the unobserved mixing proportion of a finite number of cell types are part of the latent variables to be optimized[19,21,22,23,24]
AdRoit estimates the proportions of cell types from transcriptome data of a mixed cell population including but not limited to bulk RNA-Seq and spatial transcriptome

Summary

Introduction

Bulk RNA sequencing provides the opportunity to understand biology at the whole transcriptome level without the prohibitive cost of single cell profiling. The method estimates the cell type constitution in the compound RNA-Seq samples using relevant single-cell data as a training source. AdRoit estimates the proportions of cell types from transcriptome data of a mixed cell population including but not limited to bulk RNA-Seq and spatial transcriptome.

Results

Conclusion