Integrated analysis of multimodal single-cell data

Abstract

The simultaneous measurement of multiple modalities, known as multimodal analysis, represents an exciting frontier for single-cell genomics and necessitates new computational methods that can define cellular states based on multiple data types. Here, we introduce "weighted-nearest neighbor analysis", an unsupervised framework to learn the relative utility of each data type in each cell, enabling an integrative analysis of multiple modalities. We apply our procedure to a CITE-seq dataset of hundreds of thousands of human white blood cells alongside a panel of 228 antibodies to construct a multimodal reference atlas of the circulating immune system. We demonstrate that integrative analysis substantially improves our ability to resolve cell states and validate the presence of previously unreported lymphoid subpopulations. Moreover, we demonstrate how to leverage this reference to rapidly map new datasets, and to interpret immune responses to vaccination and COVID-19. Our approach represents a broadly applicable strategy to analyze single-cell multimodal datasets, including paired measurements of RNA and chromatin state, and to look beyond the transcriptome towards a unified and multimodal definition of cellular identity. Availability: Installation instructions, documentation, tutorials, and CITE-seq datasets are available at http://www.satijalab.org/seurat

Multi-omic profiling reveals widespread dysregulation of innate immunity and hematopoiesis in COVID-19

“…2B). To prioritize downstream analysis of these cell subsets we calculated a perturbation score (30,31) for each cell type from each COVID-19 sample relative to healthy controls (see Methods). The perturbation score for each cell type is calculated by first identifying genes that display evidence of differential expression between COVID-19 samples and healthy controls, calculating the difference of pseudobulk expression vectors of these genes between COVID-19 samples and healthy controls, and finally projecting the whole transcriptome of each donor onto this vector.…”

Sfaira accelerates data and model reuse in single cell genomics

“…The mechanism introduced here to accumulate large reference databases and to fit models on extremely diverse data set collections, provides a gateway to regularization through data and to mechanistic models. In contrast to query-to-reference analysis, the models presented here can be leveraged for unconstrained data exploration 19,20 . Lastly, our framework is open for the contribution of single-cell centric models that do not primarily serve the purpose of single-cell RNA-seq embedding or cell type prediction.…”

SPINT2 controls SARS-CoV-2 viral infection and is associated to disease severity

“…As we wanted to differentiate between permissivity and infection signatures, we first looked for differentially expressed genes in SARS-CoV-2 permissive vs non-permissive cell lines and then we removed all the genes which were up- or down-regulated during infection ( Figure 2A ). We performed Differential Expression Analysis using Seurat 65 (DEA) between Calu-3 and H1299 cells in non-infected mock cells at 4 hours of culture ( z ). Then, we obtained DEGs of Calu-3 infected vs mock at 12 hours post infection ( x ); we did the same with H1299 infected cells vs mock-infected cells at 4 hpi ( y ) In all DEA we set a Log Fold Change (FC)=0.25 threshold.…”