@article{10.1093/gigascience/giz105,
    author = {Alessandrì, Luca and Cordero, Francesca and Beccuti, Marco and Arigoni, Maddalena and Olivero, Martina and Romano, Greta and Rabellino, Sergio and Licheri, Nicola and De Libero, Gennaro and Pace, Luigia and Calogero, Raffaele A},
    title = "{rCASC: reproducible classification analysis of single-cell sequencing data}",
    journal = {GigaScience},
    volume = {8},
    number = {9},
    year = {2019},
    month = {09},
    abstract = "{Single-cell RNA sequencing is essential for investigating cellular heterogeneity and highlighting cell subpopulation-specific signatures. Single-cell sequencing applications have spread from conventional RNA sequencing to epigenomics, e.g., ATAC-seq. Many related algorithms and tools have been developed, but few computational workflows provide analysis flexibility while also achieving functional (i.e., information about the data and the tools used are saved as metadata) and computational reproducibility (i.e., a real image of the computational environment used to generate the data is stored) through a user-friendly environment.rCASC is a modular workflow providing an integrated analysis environment (from count generation to cell subpopulation identification) exploiting Docker containerization to achieve both functional and computational reproducibility in data analysis. Hence, rCASC provides preprocessing tools to remove low-quality cells and/or specific bias, e.g., cell cycle. Subpopulation discovery can instead be achieved using different clustering techniques based on different distance metrics. Cluster quality is then estimated through the new metric "cell stability score" (CSS), which describes the stability of a cell in a cluster as a consequence of a perturbation induced by removing a random set of cells from the cell population. CSS provides better cluster robustness information than the silhouette metric. Moreover, rCASC's tools can identify cluster-specific gene signatures.rCASC is a modular workflow with new features that could help researchers define cell subpopulations and detect subpopulation-specific markers. It uses Docker for ease of installation and to achieve a computation-reproducible analysis. A Java GUI is provided to welcome users without computational skills in R.}",
    issn = {2047-217X},
    doi = {10.1093/gigascience/giz105},
    url = {https://doi.org/10.1093/gigascience/giz105},
    eprint = {http://oup.prod.sis.lan/gigascience/article-pdf/8/9/giz105/29808931/giz105.pdf},
}