ShortCake: An integrated platform for efficient and reproducible single-cell analysis

Single-cell analysis is a powerful method for exploring cellular heterogeneity, lineage, and spatial information by profiling the transcriptome and epigenome of individual cells (Lareau et al. (2019); Stuart et al. (2019b)). The landscape of single-cell analysis is rapidly expanding. Currently, even for single-cell RNA sequencing (scRNA-seq) alone, more than 1,800 tools are registered in the scRNA-tools database (Zappia et al. (2018)). As a result, it has become common to combine multiple tools within a single project, or to test and compare different methods for the same analytical step, while working in multiple programming languages. Managing these heterogeneous environments can pose significant challenges, especially for researchers without a bioinformatics background.

Several pipelines have been developed to facilitate such integrated single-cell data analysis using multiple tools. For instance, the Seurat ecosystem in R and the scverse in Python offer valuable frameworks for single-cell analysis (Hao et al. (2024); Virshup et al. (2023)). However, these ecosystems do not provide environments for both R and Python, and do not solve installation problems. Users often struggle with dependency conflicts, version mismatches, and complex setup procedures when trying to combine various tools for their own custom analysis workflows. Although package managers like Conda (https://anaconda.org/) are helpful, they still have difficulty reconciling conflicting package dependencies. These problems can prevent researchers from carrying out their research ideas.

Another challenge is reproducibility, especially across different computational environments. For instance, identical workflows have produced different sets of detected transcripts on macOS versus Linux (Di Tommaso et al. (2017)). Improving the reproducibility of previous studies has been a key challenge.

To address these challenges, we developed ShortCake, a Docker-based platform that consolidates various single-cell RNA-seq and ATAC-seq (scATAC-seq) analysis tools in both R and Python environments, as well as correlated command-line tools. ShortCake separates conflicting Python packages into separate virtual environments, each of which is accessible via a Jupyter Notebook kernel. This streamlines the process of switching between tools without exiting an interactive session. ShortCake also provides multiple “flavors” of the image, allowing users to download only the necessary components and conserve computational resources. Additionally, since the ShortCake Dockerfile is publicly available, users can also edit it to build custom images that include any necessary additional tools. This architecture dramatically reduces installation costs and facilitates reproducibility among users and host computers. It lowers the barrier to entry for researchers and promotes reproducibility through standardized environments.

ShortCake has been continuously updated since its inception in 2022 and has been utilized in several studies (Nagai et al. (2023); Shibata et al. (2024)). Its Docker-based approach enables users to replicate an identical analysis environment on any local machine with minimal effort, ensuring reproducible workflows. When Docker privileges are unavailable (e.g., on shared cluster servers), the same image can be run with Singularity (Kurtzer et al. (2017)) instead.

Community initiatives, such as nf-core (Marques de Almeida et al. ), have successfully standardized single-cell workflows. These initiatives provide curated Nextflow pipelines that can be executed reproducibly. This type of framework is ideal for large consortia that must run a single, agreed-upon pipeline on numerous samples in a tightly controlled environment. Conversely, ShortCake is designed for individual researchers or small groups exploring suitable case-specific workflows by iterating over alternative tools, parameter settings, and custom scripts. Rather than enforcing a fixed workflow, ShortCake offers a flexible interface that accelerates the prototyping phase for specific biological questions.

As the single-cell research field continues to advance, new technologies will emerge, such as spatial and perturbation analyses. We aim to keep up with these advances by providing the research community with well-validated tools that will facilitate their work.

No competing interest is declared.

R.N. developed ShortCake. R.N. and L.A.E.N. maintained and tested it. R.N. and L.A.E.N. wrote and revised the manuscript.

This work was supported by a Grant-in-Aid for Scientific Research under grant number 23H02466, the Japan Agency for Medical Research and Development under grant number JP23gm6310012h0004, and the JST FOREST Program under grant number JPMJFR224Y.