Current computational workflows for comparative analyses of single-cell datasets typically use discrete clusters as input when testing for differential abundance among experimental conditions. However, clusters do not always provide the appropriate resolution and cannot capture continuous trajectories. Here we present Milo, a scalable statistical framework that performs differential abundance testing by assigning cells to partially overlapping neighborhoods on a k-nearest neighbor graph. Using simulations and single-cell RNA sequencing (scRNA-seq) data, we show that Milo can identify perturbations that are obscured by discretizing cells into clusters, that it maintains false discovery rate control across batch effects and that it outperforms alternative differential abundance testing strategies. Milo identifies the decline of a fate-biased epithelial precursor in the aging mouse thymus and identifies perturbations to multiple lineages in human cirrhotic liver. As Milo is based on a cell-cell similarity structure, it might also be applicable to single-cell data other than scRNA-seq. Milo is provided as an open-source R software package at https://github.com/MarioniLab/miloR .
We thank S. Ghazanfar for feedback on the method; N. Kumasaka for comments on the manuscript; C. Suo, V. Kedlian, R. Elmentaite, J. P. Pett, K. Tuong and B. Stewart for feedback on the software package; and D. Burkhardt, M. Luecken and W. Lewis for discussions on benchmarking. J.C.M. acknowledges core funding from the European Molecular Biology Laboratory and core funding from Cancer Research UK (C9545/A29580), which supports M.D.M. E.D. and S.A.T. acknowledge Wellcome Sanger core funding (WT206194). N.C.H. is supported by a Wellcome Trust Senior Research Fellowship in Clinical Science (ref. 219542/Z/19/Z), the Medical Research Council and a Chan Zuckerberg Initiative Seed Network Grant.
- Cluster Analysis
- Gene Expression Profiling
- Sequence Analysis, RNA
- Single-Cell Analysis