Abstract
Transcriptome deconvolution aims to estimate the cellular composition of an RNA sample from its gene expression data, which in turn can be used to correct for composition differences across samples. The human brain is unique in its transcriptomic diversity, and comprises a complex mixture of cell-types, including transcriptionally similar subtypes of neurons. Here, we carry out a comprehensive evaluation of deconvolution methods for human brain transcriptome data, and assess the tissue-specificity of our key observations by comparison with human pancreas and heart. We evaluate eight transcriptome deconvolution approaches and nine cell-type signatures, testing the accuracy of deconvolution using in silico mixtures of single-cell RNA-seq data, RNA mixtures, as well as nearly 2000 human brain samples. Our results identify the main factors that drive deconvolution accuracy for brain data, and highlight the importance of biological factors influencing cell-type signatures, such as brain region and in vitro cell culturing.
Highlights
Transcriptome deconvolution aims to estimate the cellular composition of an RNA sample from its gene expression data, which in turn can be used to correct for composition differences across samples
For the partial deconvolution methods, we evaluate the effects of combining them with nine brain cell-type signature datasets that differ in biological properties or technical factors affecting RNA sequencing
We selected: CIBERSORT (CIB), a highly cited deconvolution method initially optimised for immune celltypes[18]; DeconRNASeq (DRS)[33], which implements the nonnegative least-squares approach employed by the PsychENCODE consortium[15]; MuSiC (MUS)[34], which is a single-cell-based deconvolution approach accounting for individual- and cellspecific expression variability in the signature; and dtangle (DTA)[35]
Summary
Transcriptome deconvolution aims to estimate the cellular composition of an RNA sample from its gene expression data, which in turn can be used to correct for composition differences across samples. To circumvent the confounding effect of cellular composition, gene expression measurements could in principle be carried out by experimentally isolating individual cell-types by laser capture micro-dissection[8,9], cell sorting[10–12], or single-cell and singlenucleus RNA-seq (scRNA-seq and snRNA-seq, respectively)[13]. These approaches are limited in feasibility and cost effectiveness for human brain transcriptome studies that require large sample sizes (hundreds to thousands of samples), such as eQTL studies or gene expression studies aiming to identify lowmagnitude changes in disease samples. To serve as useful covariates, cell-type composition estimates need to discriminate small differences in cellular composition[3]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
