MOTIVATION: In single-cell studies, a common question is whether there is a change in cell composition between conditions. While ideally, one needs absolute cell counts (number of cells per volumetric unit in a sample) to address these questions, current experimentation typically obtains cell counts that only carry relative information. Therefore, one should account for the compositional nature of cell count data in the statistical analysis. While recently developed methods address compositionality using compositional transformations together with a bias correction, they do not account for the uncertainty involved in estimation of the bias term, nor do they accommodate the mean-variance structure of the counts. RESULTS: Here, we introduce a statistical method, voomCLR, for assessing differences in cell composition between conditions incorporating both uncertainty on the bias term as well as acknowledging the mean-variance structure of the transformed data, by leveraging developments from the differential gene expression literature. We demonstrate the performances of voomCLR, illustrate the benefit of all components, and compare the methodology to the state-of-the-art on simulated and real single-cell gene expression datasets. AVAILABILITY AND IMPLEMENTATION: voomCLR software is available as an open-source R package on GitHub at https://github.com/johnsonandjohnson/voomCLR.

BACKGROUND: Single-cell transcriptome sequencing (scRNA-Seq) has allowed new types of investigations at unprecedented levels of resolution. Among the primary goals of scRNA-Seq is the classification of cells into distinct types. Many approaches build on existing clustering literature to develop tools specific to single-cell. However, almost all of these methods rely on heuristics or user-supplied parameters to control the number of clusters. This affects both the resolution of the clusters within the original dataset as well as their replicability across datasets. While many recommendations exist, in general, there is little assurance that any given set of parameters will represent an optimal choice in the trade-off between cluster resolution and replicability. For instance, another set of parameters may result in more clusters that are also more replicable. RESULTS: Here, we propose Dune, a new method for optimizing the trade-off between the resolution of the clusters and their replicability. Our method takes as input a set of clustering results-or partitions-on a single dataset and iteratively merges clusters within each partitions in order to maximize their concordance between partitions. As demonstrated on multiple datasets from different platforms, Dune outperforms existing techniques, that rely on hierarchical merging for reducing the number of clusters, in terms of replicability of the resultant merged clusters as well as concordance with ground truth. Dune is available as an R package on Bioconductor: https://www.bioconductor.org/packages/release/bioc/html/Dune.html . CONCLUSIONS: Cluster refinement by Dune helps improve the robustness of any clustering analysis and reduces the reliance on tuning parameters. This method provides an objective approach for borrowing information across multiple clusterings to generate replicable clusters most likely to represent common biological features across multiple datasets.