Case-control analysis of single-cell RNA-seq studies

Viktor Petukhov; Anna A. Igolkina; Rasmus Rydbirk; Shenglin Mei; Lars Christoffersen; Konstantin V. Khodosevich; Peter V. Kharchenko

doi:10.1101/2022.03.15.484475

bioRxivMar 15, 2022BioinformaticsPreprint

Case-control analysis of single-cell RNA-seq studies

Viktor Petukhov iD, Anna A. Igolkina iD, Rasmus Rydbirk iD, Shenglin Mei iD, Lars Christoffersen iD, Konstantin V. Khodosevich iD, Peter V. Kharchenko iD

PDF DOI

Structured summary available · Jump to summary ↓

Abstract

Single-cell RNA-seq (scRNA-seq) assays are being increasingly utilized to investigate specific hypotheses in both basic biology and clinically-applied studies. The design of most such studies can be often reduced to a comparison between two or more groups of samples, such as disease cases and healthy controls, or treatment and placebo. Comparative analysis between groups of scRNA-seq samples brings additional statistical considerations, and currently there is a lack of tools to address this common scenario. Based on our experience with comparative designs, here we present a computational suite (Cacoa - case-control analysis) to carry out statistical tests, exploration, and visualization of scRNA-seq sample cohorts. Using multiple example datasets, we demonstrate how application of these techniques can provide additional insights, and avoid issues stemming from inter-individual variability, limited sample size, and high dimensionality of the data.

Abstract summary

Claude Opus 5.5 · xhigh effort · 18 Oct 2026 · from the abstract

Cacoa packages statistical tests, exploration and plots for comparing groups of scRNA-seq samples, aimed at inter-individual variability, small cohorts and high-dimensional data.

Most single-cell RNA sequencing (scRNA-seq) studies come down to contrasting groups of samples - cases against controls, treated against placebo - and the authors say this kind of comparison brings additional statistical considerations that current tools do not address. Cacoa is offered as that missing layer: tests, exploration and visualization built around sample cohorts.

Key results

ClaimedA software suite, Cacoa, implements statistical testing, exploratory analysis and visualization for cohorts of scRNA-seq samples split into comparison groups.
Presented as a computational suite for case-control designs; the abstract names no individual test, model or visualization, and no software environment or availability is stated. · Abstract
ClaimedApplied to several example datasets, the approach is said to add findings and to sidestep pitfalls of group comparisons: variability between donors, few samples per group, and the dimensionality of the data.
Asserted from multiple example datasets, none of which is named; no effect sizes, false-positive rates or comparisons with alternative differential testing strategies are reported. · Abstract

Methods, data and reproducibility

A computational suite for case-control designs in scRNA-seq: it treats the study as a comparison between groups of samples and provides statistical tests plus exploratory and visual summaries at that level, illustrated on several example cohorts.

several example scRNA-seq sample cohorts with case-control or comparable group structure, used for demonstration

Methods & tools: Cacoa

Limitations (2)

Neither the statistical methods nor the example datasets are identified in the available text, so what Cacoa does and on what data cannot be judged from it. · noted in this summary
No benchmark, simulation or false-positive assessment is reported against existing ways of testing differences between groups of single-cell samples. · noted in this summary

Read this if you analyse scRNA-seq cohorts with case and control groups and need sample-level statistics rather than per-cell tests.