Dictionary learning for integrative, multimodal and scalable single-cell analysis.

Mapping single-cell sequencing profiles to comprehensive reference datasets provides a powerful alternative to unsupervised analysis. However, most reference datasets are constructed from single-cell RNA-sequencing data and cannot be used to annotate datasets that do not measure gene expression. Here we introduce 'bridge integration', a method to integrate single-cell datasets across modalities using a multiomic dataset as a molecular bridge. Each cell in the multiomic dataset constitutes an element in a 'dictionary', which is used to reconstruct unimodal datasets and transform them into a shared space. Our procedure accurately integrates transcriptomic data with independent single-cell measurements of chromatin accessibility, histone modifications, DNA methylation and protein levels. Moreover, we demonstrate how dictionary learning can be combined with sketching techniques to improve computational scalability and harmonize 8.6 million human immune cell profiles from sequencing and mass cytometry experiments. Our approach, implemented in version 5 of our Seurat toolkit ( http://www.satijalab.org/seurat ), broadens the utility of single-cell reference datasets and facilitates comparisons across diverse molecular modalities.

Diverse macrophage populations contribute to distinct manifestations of human cutaneous graft-versus-host disease.

BACKGROUND: Graft-versus-host disease (GvHD) is a major life-threatening complication of allogeneic haematopoietic stem cell transplantation (HSCT), limiting the broad application of HSCT for haematological malignancies. Cutaneous GvHD is described as a post-transplant inflammatory reaction by skin-infiltrating donor T cells and remaining recipient tissue-resident memory T cells. Despite the major influence of lymphocytes on GvHD pathogenesis, the complex role of mononuclear phagocytes (MNPs) in tissues affected by GvHD is increasingly appreciated. OBJECTIVES: To characterize the identity, origin and functions of MNPs in patients with acute cutaneous GvHD. METHODS: Using single-cell RNA sequencing and multiplex tissue immunofluorescence, we identified an increased abundance of MNPs in skin and blood from 36 patients with acute cutaneous GvHD. In cases of sex-mismatched transplantation, we used expression of X-linked genes to detect rapid tissue adaptation of newly recruited donor MNPs resulting in similar transcriptional states of host- and donor-derived macrophages within GvHD skin lesions. RESULTS: We showed that cutaneous GvHD lesions harbour expanded CD163+ tissue-resident macrophage populations with anti-inflammatory and tissue-remodelling properties including interleukin-10 cytokine production. Cell-cell interaction analyses revealed putative signalling to strengthen regulatory T-cell responses. Notably, macrophage polarization in chronic cutaneous GvHD types was proinflammatory and drastically differed from acute GvHD, supporting the notion of distinct cellular players in different clinical GvHD subtypes. CONCLUSIONS: Overall, our data reveal a surprisingly dynamic role of MNPs after HSCT. Specific and time-resolved targeting to repolarize this cell subset may present a promising therapeutic strategy in combatting GvHD skin inflammation.

Single-cell chromatin state analysis with Signac.

The recent development of experimental methods for measuring chromatin state at single-cell resolution has created a need for computational tools capable of analyzing these datasets. Here we developed Signac, a comprehensive toolkit for the analysis of single-cell chromatin data. Signac enables an end-to-end analysis of single-cell chromatin data, including peak calling, quantification, quality control, dimension reduction, clustering, integration with single-cell gene expression datasets, DNA motif analysis and interactive visualization. Through its seamless compatibility with the Seurat package, Signac facilitates the analysis of diverse multimodal single-cell chromatin data, including datasets that co-assay DNA accessibility with gene expression, protein abundance and mitochondrial genotype. We demonstrate scaling of the Signac framework to analyze datasets containing over 700,000 cells.