Comprehensive evaluation of human brain gene expression deconvolution methods

Abstract

Gene expression measurements, similar to DNA methylation and proteomic measurements, are influenced by the cellular composition of the sample analysed. Deconvolution of bulk transcriptome data aims to estimate the cellular composition of a sample from its gene expression data, which in turn can be used to correct for composition differences across samples. Although a multitude of deconvolution methods have been developed, it is unclear whether their performance is consistent across tissues with different complexities of cellular composition. The human brain is unique in its transcriptomic diversity, expressing the highest diversity of alternative splicing isoforms and non-coding RNAs. It comprises a complex mixture of cell-types including transcriptionally similar sub-types of neurons, which undergo gene expression changes in response to neuronal activity. However, a comprehensive assessment of the accuracy of transcriptome deconvolution methods on human brain data is currently lacking. Here we carry out the first comprehensive comparative evaluation of the accuracy of deconvolution methods for human brain transcriptome data, and assess the tissue-specificity of our key observations by comparison with transcriptome data from human pancreas and heart. We evaluate 8 transcriptome deconvolution approaches, covering all main classes: 4 partial deconvolution methods, each applied with 9 different cell-type signatures, 2 enrichment methods, and 2 complete deconvolution methods. We test the accuracy of cell-type estimates using in silico mixtures of single-cell RNA-seq data, mixtures of neuronal and glial RNA, as well as nearly 2,000 human brain samples. Our results bring several important insights into the performance of transcriptome deconvolution: (a) We find that cell-type signature data has a stronger impact on brain deconvolution accuracy than the choice of method. (b) We demonstrate that biological factors influencing brain cell-type signature data (e.g. brain region, in vitro cell culturing), have stronger effects on the deconvolution outcome than technical factors (e.g. RNA sequencing platform). (c) We find that partial deconvolution methods outperform complete deconvolution methods on human brain data. To facilitate wider implementation of correction for cellular composition, we develop a webtool that implements the best performing methods, and is available at https://voineagulab.shinyapps.io/BrainDeconvShiny/ .