The Growing Toolkit for Single-Cell RNA Sequencing Analysis

Single-cell RNA sequencing (scRNA-seq) provides unprecedented detail into complex cellular processes, with each experiment generating millions of reads that capture the transcriptional activity of millions of individual cells. This capability makes scRNA-seq powerful for studying complex systems, but also challenges researchers to translate the data into clear biological conclusions. To meet these analytical demands, robust computational tools have become essential for tasks such as identifying cell types, comparing conditions, correcting for batch effects, and distinguishing genuine biology from technical noise. While early work often depended on custom code, the field now benefits from widely used frameworks and specialized packages that provide standardized workflows. This article reviews several influential platforms and highlights the capabilities and advances of each one.

Seurat
Early scRNA-seq studies provided comprehensive gene expression profiles but lost information about the location of cells in tissues, since dissociation removed them from their native environment. Other methods, such as in situ hybridization or immunohistochemistry, preserved spatial context but could only track a few genes at once, thereby limiting discovery. To address this gap, Rahul Satija, working with Aviv Regev at the Broad Institute, developed Seurat, which integrates transcriptomic signatures with reference maps of landmark genes to infer each cell’s spatial origin.¹ Its name pays homage to Georges Seurat, whose pointillist paintings resemble the way individual transcript measurements combine into cellular patterns.

Since its release in 2015, Seurat has become one of the most widely recognized toolkits for scRNA-seq analysis. It has grown into a versatile platform that supports core workflows such as quality control, clustering, dimensional reduction, and marker identification, and it enables integration across datasets, technologies, and even species. Along with transcriptomics, Seurat also supports multimodal profiling by combining RNA with chromatin accessibility, protein, or perturbation data.

Seurat has been cited in thousands of publications and has served an important role in many high-profile projects, including large-scale efforts such as the Human Cell Atlas.² Benchmarking studies have shown that Seurat is consistently among the top performers for batch correction, clustering accuracy, and multiomic integration.^3-5 It is also widely recommended as a first-choice tool for single-cell RNA-seq because of its effective quality control, simple processing, and strong scalability for large datasets.⁶

Part of Seurat’s popularity comes from its extensive learning resources. The Satija Lab at the New York Genome Center continues to maintain and expand Seurat while providing tutorials, example datasets, and vignettes that guide users step by step through common analyses. It is also widely taught in workshops and university courses, where it often serves as the default introduction to computational single-cell methods. A large and active community also contributes code, extensions, and troubleshooting support through GitHub, forums, and online tutorials.
The latest release, Seurat v5, introduces bridge integration for combining modalities such as scRNA-seq and scATAC-seq, streamlined workflows for comparing integration strategies, and infrastructure for analyzing datasets with millions of cells through sketch-based and disk-backed approaches powered by the BPCells package.⁷ Version five also expands support for sequencing- and imaging-based spatial data, enabling analyses such as deconvolution and niche identification. These new features show Seurat’s ability to keep pace with an increasingly multimodal field. With regular updates and broader integration of spatial and epigenetic data, it is likely to remain a central toolkit for single-cell analysis.


Trailmaker
Conversations with single-cell researchers revealed that data analysis was a major bottleneck, said Vicky Morrison, Ph.D., Senior Product Manager for Software at Parse Biosciences. “Bench scientists were sitting on data for months because they lacked the programming skills to analyze the data themselves using common analysis methods that require coding in Python or R,” she explained. At the same time, bioinformaticians were overloaded with projects while guiding colleagues through analyses. These challenges inspired the development of Trailmaker.

“We built Trailmaker to empower wet-lab scientists to bring their deep understanding of their own systems directly into the data analysis process by removing the barrier of having to write code,” Morrison stated. The plan was to make analysis intuitive and accessible, while also supporting collaboration with bioinformaticians through a shared platform.

Designed as an end-to-end solution, Trailmaker takes researchers from raw FASTQ files all the way to publication-ready figures. It does this using two main modules, Pipeline and Insights. Pipeline processes FASTQ files generated by Parse’s Evercode technology, while Insights supports downstream analysis of count matrices in a technology-agnostic manner. Morrison emphasized that “the platform incorporates common analysis methods that are considered gold standard in the single cell field, including the Seurat and Scanpy workflows, pseudo-bulk differential expression analysis, and trajectory analysis.” By packaging these workflows, Trailmaker makes them accessible to researchers who may not be familiar with the coding frameworks typically required to run them.

The platform builds on this accessibility with an intuitive interface, detailed documentation, and free educational resources, supported by webinars, courses, and technical assistance. Another strength is its integrated environment, which manages the entire pipeline without forcing researchers to piece together fragmented tools. “This lowers the barrier of entry, making it easy to get started with Parse Biosciences technology and with single-cell in general,” stated Morrison. Trailmaker is also designed for flexibility. Experienced users can enter or exit at different stages, using its automated pipelines while still integrating community-developed tools. Morrison noted that the goal is to let researchers analyze data in whatever way best suits their needs.

Trailmaker has already gained traction within the research community. Formerly known as Cellenics, it has been cited in over 35 publications and has more than 1,100 active users each month. One notable application comes from Nick Ciccone, Research Fellow at the University of Oxford, who used Trailmaker and Evercode to study how circadian rhythm disruption heightens liver sensitivity to stress hormones. His work revealed hepatocytes as key mediators of fat accumulation and disease, suggesting therapeutic strategies targeting glucocorticoid signaling.

Morrison also noted that Trailmaker is updated regularly with new features and performance improvements. In June, Parse added functionality that allows researchers to upload count matrices from multiple technologies into a single project, enabling side-by-side analysis within one integrated view. Another update expanded the Pipeline module to support Evercode TCR and BCR immune profiling kits. That update brings immune repertoire analysis into the platform. “We are and will continue to enhance Trailmaker with new features and improvements, including immune profiling data visualization, guided by the feedback we receive from our users,” emphasized Morrison.

Trailmaker aims to make single-cell analysis more accessible to researchers, while still offering flexibility for computational specialists. Its growing number of users and publications demonstrate that it fills a practical need in the single-cell community. As Morrison concluded, “Trailmaker really speaks for itself, and it’s free to explore. Simply sign in, explore your own data, or dive into one of the datasets in our repository, and see firsthand what Trailmaker can do for you.”

Scanpy
Initial versions of tools like Seurat, Monocle, and Cell Ranger supported integrated workflows for exploring cell states and dynamics, yet they lagged behind the rapidly growing scale of datasets with millions of cells. This scalability bottleneck restricted how effectively researchers could analyze large datasets. Recognizing these limitations, Alex Wolf, Philipp Angerer, and Fabian Theis developed Scanpy, a Python-based framework that scales efficiently to datasets with millions of cells and integrates seamlessly with machine learning libraries.⁸

The core workflow of Scanpy is built around the AnnData data structure, which stores gene expression matrices with detailed annotations. Its pipeline includes preprocessing steps such as normalization, confounder regression, and identification of highly variable genes. For dimensionality reduction and visualization, it offers PCA, t-SNE, diffusion maps, and graph-based layouts. Clustering is performed with algorithms like Louvain, alongside tools for identifying marker genes through differential expression tests. Scanpy also supports pseudotime and trajectory inference with diffusion pseudotime, enabling the reconstruction of branching cell lineages.

Scanpy has grown through a series of major updates that steadily expanded its speed, efficiency, and analytical scope.⁹ Early updates introduced UMAP, the Neighbors class, and Leiden clustering, while also improving batch correction and quality control tools. Later versions added spatial data support, refined highly variable gene selection, and strengthened differential expression analysis. More recent updates introduced Pearson Residuals, improved package compatibility, and fixed long-standing issues in HVG functions. Over time, Scanpy has also gained numerous extensions, including tools for manual cell selection, immune repertoire analysis, cell fate mapping, and many more.^10-12

The Scanpy website offers helpful documentation that provides tutorials that guide users through preprocessing, clustering, visualization, and advanced workflows. It also includes a comprehensive API reference, practical “how-to” guides, and detailed release notes that track new features and changes. In addition, users can find installation instructions and links to the community for questions and contributions, making the site a complete resource for learning and applying Scanpy.

This framework has become a central resource in the field that has allowed researchers to analyze ever-larger datasets while adapting to new experimental modalities. Scanpy’s influence now extends across transcriptomic, spatial, and epigenomic studies, and ongoing contributions from developers and users highlight its role as an actively developing platform for the single-cell community.

The Growing Ecosystem
In addition to these core platforms, many other scRNA-seq analysis tools and frameworks are available. Commercial examples include DNAnexus, Basepair, and ROSALIND, along with open-source packages such as Monocle, scVI, and STAR. Although an exhaustive list is not possible, the scRNA-tools database catalogs hundreds of single-cell packages, giving researchers a central resource for identifying methods that match their experimental needs. This expanding ecosystem reflects how scRNA-seq, combined with computational innovation, is advancing our understanding of complex biology.

References

1. Satija R, Farrell JA, Gennert D, Schier AF, Regev A. Spatial reconstruction of single-cell gene expression data. Nat Biotechnol. 2015;33(5):495-502. doi:10.1038/nbt.3192
2. Lake BB, Menon R, Winfree S, et al. An atlas of healthy and injured cell states and niches in the human kidney. Nature. 2023;619(7970):585-594. doi:10.1038/s41586-023-05769-3
3. Tran HTN, Ang KS, Chevrier M, et al. A benchmark of batch-effect correction methods for single-cell RNA sequencing data. Genome Biol. 2020;21(1):12. Published 2020 Jan 16. doi:10.1186/s13059-019-1850-9
4. Xiao C, Chen Y, Meng Q, Wei L, Zhang X. Benchmarking multi-omics integration algorithms across single-cell RNA and ATAC data. Brief Bioinform. 2024;25(2):bbae095. doi:10.1093/bib/bbae095
5. Zhang S, Li X, Lin J, Lin Q, Wong KC. Review of single-cell RNA-seq data clustering for cell-type identification and characterization. RNA. 2023;29(5):517-530. doi:10.1261/rna.078965.121
6. Lee MYY, Kaestner KH, Li M. Benchmarking algorithms for joint integration of unpaired and paired single-cell RNA-seq and ATAC-seq data. Genome Biol. 2023;24(1):244. Published 2023 Oct 24. doi:10.1186/s13059-023-03073-x
7. Hao Y, Stuart T, Kowalski MH, et al. Dictionary learning for integrative, multimodal and scalable single-cell analysis. Nat Biotechnol. 2024;42(2):293-304. doi:10.1038/s41587-023-01767-y
8. Wolf FA, Angerer P, Theis FJ. SCANPY: large-scale single-cell gene expression data analysis. Genome Biol. 2018;19(1):15. Published 2018 Feb 6. doi:10.1186/s13059-017-1382-0
9. Rich JM, Moses L, Einarsson PH, et al. The impact of package selection and versioning on single-cell RNA-seq analysis. Preprint. bioRxiv. 2024;2024.04.04.588111. Published 2024 Apr 11. doi:10.1101/2024.04.04.588111
10. Dedden M, Wiendl M, Müller TM, Neurath MF, Zundler S. Manual cell selection in single cell transcriptomics using scSELpy supports the analysis of immune cell subsets. Front Immunol. 2023;14:1027346. Published 2023 Apr 25. doi:10.3389/fimmu.2023.1027346
11. Sturm G, Szabo T, Fotakis G, et al. Scirpy: a Scanpy extension for analyzing single-cell T-cell receptor-sequencing data. Bioinformatics. 2020;36(18):4817-4818. doi:10.1093/bioinformatics/btaa611
12. Lange M, Bergen V, Klein M, et al. CellRank for directed single-cell fate mapping. Nat Methods. 2022;19(2):159-170. doi:10.1038/s41592-021-01346-6