A simple, scalable approach to building a cross-platform transcriptome atlas

Abstract

Gene expression atlases have transformed our understanding of the development, composition and function of human tissues. New technologies promise improved cellular or molecular resolution, and have led to the identification of new cell types, or better defined cell states. But as new technologies emerge, information derived on old platforms becomes obsolete. We demonstrate that it is possible to combine a large number of different profiling experiments summarised from dozens of laboratories and representing hundreds of donors, to create an integrated molecular map of human tissue. As an example, we combine 850 samples from 38 platforms to build an integrated atlas of human blood cells. We achieve robust and unbiased cell type clustering using a variance partitioning method, selecting genes with low platform bias relative to biological variation. Other than an initial rescaling, no other transformation to the primary data is applied through batch correction or renormalisation. Additional data, including single-cell datasets, can be projected for comparison, classification and annotation. The resulting atlas provides a multi-scaled approach to visualise and analyse the relationships between sets of genes and blood cell lineages, including the maturation and activation of leukocytes in vivo and in vitro. In allowing for data integration across hundreds of studies, we address a key reproduciblity challenge which is faced by any new technology. This allows us to draw on the deep phenotypes and functional annotations that accompany traditional profiling methods, and provide important context to the high cellular resolution of single cell profiling. Here, we have implemented the blood atlas in the open access Stemformatics.org platform, drawing on its extensive collection of curated transcriptome data. The method is simple, scalable and amenable for rapid deployment in other biological systems or computational workflows.

Rank-in: enabling integrative analysis across microarray and RNA-seq for cancer

“…Rank-In was comprehensively validated on both SEQC cell data and TCGA clinical data, where the same biological samples were tested on both microarray and RNA-seq platforms. The performance of Rank-In was evaluated via the unsupervised hierarchical clustering effects and the ability to pick up true differential expression genes (DEGs), in comparing with four representative peers, including uncorrected method without processing, Combat ( 12 ), SVA ( 15 ) and Angel’s method ( 21 ).…”

“…Yet on clinical samples of both GBM and colon cancer, its performance shows fluctuation, with general achievement better than Combat, but worse than SVA. This may be related to the inherent limitation of the ‘marker genes’ strategy, where the personal variation in health tissues might be not enough to become the ‘prominent features’ in data collection ( 21 ). While in Rank-In, both relative and absolute intensity of gene expression are considered for all overlapping genes between datasets.…”

“…Currently, SVA and its updated version ( 16 ) have been widely applied to bulk and single-cell RNA-seq analysis ( 17 , 18 ). Meanwhile, spike-in marker genes were also adopted to adjust cross-platform biases, such as housekeeping genes ( 19 , 20 ) or so-called ‘bias-low gene sets’ ( 21 ). Yet this hypothesis has been questionable ( 22 ), as the filtered gene signatures were reported to be highly unstable ( 23 ), depending on patient samples, disease states ( 24 ) or tissues ( 25 ).…”

An integrated analysis of human myeloid cells identifies gaps in in vitro models of in vivo biology

“…Atlas construction was developed as described in ( Angel et al., 2020 ) and is composed of 44 datasets, 901 samples, and 3,757 genes. Samples can be colored by cell type, sample source, progenitor type, tissue, disease, activation status, or dataset.…”

An Unbiased Machine Learning Exploration Reveals Gene Sets Predictive of Allograft Tolerance After Kidney Transplantation

“…Although prior meta-analysis investigations have sought to identify key biomarkers for tolerance (3,(5)(6)(7)(8)12), the potential influence of the diverse RNA sequencing platforms used across the existing studies has remained a confounding variable. Several published methods address the problem of integrating data across platforms (13,14). Nevertheless, when sequencing platforms vary, the sample management and gene expression profiles can differ substantially.…”