Detection of differentially abundant cell subpopulations in scRNA-seq data

Abstract

Comprehensive and accurate comparisons of transcriptomic distributions of cells from samples taken from two different biological states, such as healthy versus diseased individuals, are an emerging challenge in single-cell RNA sequencing (scRNA-seq) analysis. Current methods for detecting differentially abundant (DA) subpopulations between samples rely heavily on initial clustering of all cells in both samples. Often, this clustering step is inadequate since the DA subpopulations may not align with a clear cluster structure, and important differences between the two biological states can be missed. Here, we introduce DA-seq, a targeted approach for identifying DA subpopulations not restricted to clusters. DA-seq is a multiscale method that quantifies a local DA measure for each cell, which is computed from its k nearest neighboring cells across a range of k values. Based on this measure, DA-seq delineates contiguous significant DA subpopulations in the transcriptomic space. We apply DA-seq to several scRNA-seq datasets and highlight its improved ability to detect differences between distinct phenotypes in severe versus mildly ill COVID-19 patients, melanomas subjected to immune checkpoint therapy comparing responders to nonresponders, embryonic development at two time points, and young versus aging brain tissue. DA-seq enabled us to detect differences between these phenotypes. Importantly, we find that DA-seq not only recovers the DA cell types as discovered in the original studies but also reveals additional DA subpopulations that were not described before. Analysis of these subpopulations yields biological insights that would otherwise be undetected using conventional computational approaches.

The transcription factor HIF-1α mediates plasticity of NKp46+ innate lymphoid cells in the gut

“…2 C). The distinct distribution of both genotypes within the ILC3 cluster was further corroborated by a differential abundance analysis based on differential expression of clusterdefining genes (Zhao et al, 2021;Fig. 2 D).…”

SOTIP: a Unified Framework for Microenvironment Modelling with Spatial Omics Data

“…1f middle and right). With this simulation, differential abundance analysis based on non-spatial scRNA-seq 40,42,43 failed to detect any difference between the two samples, but SOTIP successfully highlighted (Fig. 1g) the major differential MECNs (C1-C2 boundary in sample 1, and C1-C3 boundary in sample 2), by estimating the relative likelihood of observing each microenvironment in the two tissue samples 43 (see Methods).…”

“…disease status, drug treatment, and experimental perturbation) with high specificity. This task is in analogy with the differential abundance analysis task 40,42,43 in single cell analysis, but unavailable in spatial omics analysis. To demonstrate the utility of SOTIP on this task, and as a sanity check, we firstly applied DMA in a spatial metabolomics dataset including 2 samples from healthy and fibrotic liver (Fig.…”

“…To demonstrate SOTIP's capability, we choose the more challenging case to identify differential MECNs between mixed (immune cells mixed with tumor cells) 12 and compartmentalized (immune cells spatially separated from tumor cells) 12 samples. The challenge to distinguish the mixed from the compartmentalized came from the fact that these two subtypes of TNBC samples typically contained similar cell type compositions 12 , such that they could not be easily identified by current singlecell based analysis 40,42 , or protein co-expression analysis 66 . To this end, we chose the most representative samples, suggested by the original research 12 , from the two subtypes of TNBC, i.e., patient 4 for compartmentalized and patient 12 for mixed (Fig.…”

DCATS: differential composition analysis for complex single-cell experimental designs

“…It assumes that cell counts follow a negative binomial distribution and uses representative cells instead of all cells to improve program efficiency. DAseq [9] also makes use of KNN graph. It calculates multiscale differential abundance score by counting numbers of cells coming from different biological states while varying k. This multiscale differential abundance score is what DAseq used to infer cell states that have differential abundance.…”