Comparing tools for single-cell and spatial transcriptomics analysis

A comprehensive comparison to help you choose the right platform for your research: OmnibusX, 10X Genomics Loupe Browser, CellxGene, and BioTuring BrowserX + SpatialX

Introduction

Single-cell RNA sequencing (scRNA-seq) and spatial transcriptomics have transformed how we study cellular heterogeneity, tissue architecture, and gene regulation. Yet as these technologies have matured, so has the challenge of analyzing the data they produce. Researchers now face a fragmented landscape of software tools, each with different strengths, trade-offs, and target audiences.

In this post, we take a detailed look at 4 current platforms for single-cell and spatial data analysis: OmnibusX, 10x Genomics Loupe Browser, CellxGene, and BioTuring BBrowserX + SpatialX. We evaluate each across the dimensions that matter most in practice: supported data types and formats, analytical depth, ease of use, visualization capabilities, data privacy, collaboration features, and cost. Whether you are a bench biologist exploring your first dataset or a bioinformatician managing institutional pipelines, this guide is designed to help you identify the platform that best fits your workflow.

No tool is perfect for every use case. Our goal is not to declare a winner but to give you the information you need to make an informed decision.

At a glance: Feature comparison

OmnibusX

Overview: OmnibusX is a privacy-centric desktop application that consolidates end-to-end workflows for scRNA-seq, scATAC-seq, bulk RNA-seq, spatial transcriptomics, and multi-omics data into a single, code-free interface. Published in PLOS Computational Biology (10.1371/journal.pcbi.1013480), OmnibusX is built on a Python analytics server and an Electron/React GUI. It integrates widely adopted open-source packages - Scanpy, DESeq2, SciPy, scikit-learn, Palantir, CellPhoneDB - into structured, reproducible pipelines, while adding proprietary modules such as a cell-type prediction engine and an interactive plot editor.

Data import and format support: OmnibusX accepts a wide range of input formats. For scRNA-seq, it reads 10x Cell Ranger HDF5, Scanpy .h5ad, Seurat .rds (v3–v5), MTX+TSV/CSV bundles, and plain CSV/TSV count matrices. TCR/BCR contig annotation files from Cell Ranger are also supported. Spatial transcriptomics support covers 10x Visium, Visium HD, and Xenium, Vizgen MERFISH, NanoString GeoMx DSP, and Slide-seq, including gene expression matrices, spatial coordinate files, and H&E or immunofluorescence images (PNG, JPEG, TIFF, OME-TIFF). Upon upload, OmnibusX automatically converts data into an optimized internal structure, standardizes gene identifiers via Ensembl annotations, and preserves the original files in a read-only state.

Processing pipeline: OmnibusX provides modality-specific default pipelines designed for the unique characteristics of each data type. For scRNA-seq, the standard workflow includes interactive QC with real-time threshold visualization, filtering, optional doublet detection and HTO demultiplexing, log normalization (or CLR normalization for ADT data), highly variable gene selection, PCA, UMAP/t-SNE embedding, and clustering. For spatial data, dedicated pipelines are adapted to the output of different spatial platforms and combine gene expression analysis with image processing. Importantly, raw unfiltered data is always retained, so users can adjust thresholds and reprocess datasets at any time without re-uploading.

Cell type annotation: OmnibusX includes a proprietary cell-type prediction engine based on a curated database of 166 cell type and subtype-specific marker gene sets, derived from 280 publications and validated across multiple datasets. The algorithm builds on AUCell enrichment scoring with nearest-neighbor consensus smoothing. Automatic predictions are applied to scRNA-seq, Visium HD, Xenium, MERFISH, and Slide-seq datasets from human and mouse. Users can also supply custom marker sets or manually refine annotations. Multi-level subclustering supports iterative refinement of cell populations.

Downstream analysis: The platform provides a comprehensive suite of downstream modules: compositional analysis, differential expression (t-test, Wilcoxon for single-cell; DESeq2 for pseudobulk/bulk), gene set enrichment (GSEA via GSEApy, Fisher's exact test, AUCell), interactive heatmaps with hierarchical clustering, trajectory and pseudotime analysis via Palantir, TCR/BCR repertoire profiling (diversity, repertoire overlap, spectratyping), and cell–cell interaction analysis using CellPhoneDB for spatial datasets. Batch correction is available through Harmony and ComBat, applied to the embedding space while preserving raw values for statistical testing.

Visualization and plot customization: A distinctive feature of OmnibusX is its built-in plot-editing module. Nearly every generated visualization - scatter plots, heatmaps, bar charts, violin plots, box plots, dot plots, line plots, etc. - can be customized within the application. Users can adjust axis labels, font styles, color palettes, element ordering, marker sizes, layout settings, and more. A dedicated color palette manager allows researchers to define and reuse palettes across datasets. This allows production of publication-ready figures without exporting to external tools like Illustrator or ggplot2.

Spatial analysis: OmnibusX supports a broad range of spatial transcriptomics platforms, including 10x Visium, Visium HD, Xenium, Vizgen MERFISH, NanoString GeoMx DSP, and Slide-seq. The spatial viewer provides synchronized visualization of gene expression overlaid on high-resolution H&E-stained and immunofluorescence tissue images, with multi-resolution pyramid indexing for real-time zoom and pan on large images, channel merging and intensity adjustment for IF images, and cross-sample comparison through a unified spatial grid. Region selection is available via a built-in lasso tool. As of December 2025, OmnibusX also includes a dedicated tissue segmentation module that automatically identifies anatomical structures across a wide range of tissue types (brain, pancreas, colon, liver, ovary, kidney, lung, skin, lymph node, and more). Users can segment tissue at multiple levels of detail - from broad anatomical regions to sub-tissue compartments - enabling refined analysis of microenvironments directly within the application.

Privacy and deployment: All analytical processing occurs locally on the user's machine. Communication with the OmnibusX server is limited to authentication, license validation, and downloading reference files (gene annotations, marker sets). No input data or results are transferred externally. An enterprise edition supports centralized deployment on private servers (on-premises or private cloud), with hierarchical user roles, group-based collaboration, and fine-grained data-sharing permissions (READ, WRITE, DELETE).

Reproducibility: OmnibusX automatically records all analysis parameters, visualization settings, and workflow states, allowing users to revisit and modify previous analyses without re-running computational steps.

Pros

• End-to-end pipeline: from raw data QC all the way to publication-quality figures, without leaving the app
• Wide format support (h5ad, rds, HDF5, MTX, CSV, spatial files)
• Broad spatial platform coverage (Visium, Visium HD, Xenium, MERFISH, GeoMx, Slide-seq) with built-in tissue segmentation
• Proprietary cell-type prediction engine with curated marker sets validated across 280 publications
• Strong data privacy by design - all computation is local, no data uploaded to external servers
• Interactive plot editor with deep customization eliminates the need for external figure-editing software
• Enterprise edition enables institutional collaboration with data governance controls
• Modality-specific default pipelines reduce guesswork for non-computational users

Cons

• Not open-source
• Paid license required
• No native R integration

10x Genomics Loupe Browser

Overview: Loupe Browser is a free desktop application developed by 10x Genomics for interactive visualization and exploration of data generated by 10x Genomics platforms. It is tightly integrated with the 10x ecosystem and is designed to be the default viewer for output files produced by Cell Ranger and Space Ranger pipelines.

Data import and format support: Loupe Browser works primarily with .cloupe files, which are generated automatically by 10x Genomics' upstream processing pipelines (Cell Ranger, Space Ranger). The tool does not accept third-party formats such as .h5ad or .rds. This tight coupling means that Loupe Browser is effectively limited to data produced by 10x platforms. Loupe also supports the import of pre-computed Seurat or Scanpy results if they are first converted to .cloupe format using the LoupeR or LoupeR packages.

Processing capabilities: Loupe Browser is primarily a visualization tool, not a full data processing pipeline. It does not perform QC filtering or normalization from scratch. These steps are handled upstream by Cell Ranger or Space Ranger. However, as of version 7.0+, Loupe does include built-in differential expression analysis using a pseudo-bulk negative binomial test (the same sSeq/edgeR method used by Cell Ranger), and version 8.0 added reclustering for unlimited barcodes. Users can also manually annotate cell types and explore gene expression patterns. Still, the heavy computational lifting (QC, normalization, embedding) is expected to have been done beforehand.

Visualization and exploration: Where Loupe Browser excels is in smooth, responsive exploration of 10x datasets. It provides interactive UMAP/t-SNE scatter plots, spatial tissue views for Visium and Visium HD data, split-view modes for comparing conditions, feature expression overlays, co-expression analysis (v8.0+), and violin/box plot views. For V(D)J data, the Loupe V(D)J Browser offers clonotype exploration. The interface is polished and optimized for the 10x data format, resulting in fast rendering even for large datasets.

Spatial analysis. For Visium and Visium HD data, Loupe Browser provides an integrated spatial viewer that overlays cluster labels and gene expression onto tissue images. Visium HD support (added in v8.0) enables single-cell-scale resolution with over 11 million capture squares per slide. Users can manually select and annotate tissue regions, compare spatial distributions of gene expression, and view spot-level or cell-level data depending on the platform. The spatial experience is one of Loupe's strongest features, particularly for 10x spatial platforms.

Collaboration and sharing: Loupe Browser uses a file-based sharing model. Users can distribute .cloupe files to collaborators, who can then open them in their own Loupe Browser installation. This is simple and effective but does not include server-based collaboration, version control, or permission management.

Pricing. Loupe Browser is free to download and use.

Pros

• Completely free - no licensing fees, no trial period restrictions
• Seamless integration with Cell Ranger, Space Ranger, and other 10x pipelines
• V(D)J clonotype browser
• All data stays local on the user's machine
• Simple file-based sharing via .cloupe files makes it easy to hand off datasets to collaborators
• Low barrier to entry for labs already using 10x platforms

Cons

• Only works with 10x Genomics data, cannot import .h5ad, .rds, or other common formats without conversion
• No full data processing pipeline, no QC, normalization, or batch correction from raw data (reclustering available in v8.0+)
• Differential expression is available (pseudo-bulk, v7.0+), but no enrichment analysis, trajectory analysis, or cell–cell interaction modules
• No automated cell-type annotation
• Limited plot customization
• No built-in collaboration platform or user management

CellxGene

Overview: CellxGene is an open-source suite of tools developed by the Chan Zuckerberg Initiative (CZI) designed to help researchers discover, explore, and analyze single-cell datasets. The platform consists of two major components: CZ CellxGene Discover, a web-based portal hosting a curated collection of published single-cell datasets, and CellxGene Explorer, a lightweight local visualization tool for viewing individual .h5ad files.

Exploration and visualization: CellxGene (both hosted and local) provides interactive UMAP/t-SNE scatter plots with gene expression overlays, support for categorical and continuous metadata coloring, differential expression between user-selected cell groups (using simple statistical tests), and basic filtering and subsetting. The interface is clean and intuitive, designed for rapid exploration rather than deep analysis. Users can explore published datasets directly in the browser without downloading anything.

Processing capabilities: CellxGene is not a data processing platform. It does not perform QC, normalization, batch correction, clustering, or any upstream computational steps. The hosted datasets have already been processed by their original authors and standardized by the CZI curation team. For users who want to perform their own processing, CellxGene provides data in .h5ad format, which can be loaded into Scanpy or other Python-based analysis frameworks.

Data privacy: For the hosted version, data is stored on CZI servers and accessed through the web. Users exploring published datasets are accessing publicly available data, so privacy is generally not a concern. The local CellxGene Explorer runs entirely on the user's machine and does not send data externally. However, there is no built-in processing pipeline in the local version - it is purely a viewer.

Community and ecosystem: CellxGene benefits enormously from its open-source nature and CZI backing. It has a large and active community, extensive documentation, and deep integration with the Python single-cell ecosystem (Scanpy, AnnData).

Pricing: CellxGene is entirely free and open-source.

Pros

• Completely free and open-source
• Web-based access to curated public datasets requires zero setup - just a browser
• Clean, intuitive interface for rapid data exploration
• Excellent for cross-study meta-analysis and building reference panels
• Strong community support and active development by CZI

Cons

• Not an analysis platform, no QC, normalization, clustering, batch correction, or downstream analytics
• No multi-omics (CITE-sesq/ADT) or TCR/BCR analysis capabilities
• Limited visualization options
• Hosted version requires data to be public (or uploaded to CZI servers)
• No built-in cell-type annotation, trajectory analysis, enrichment, or cell–cell interaction modules
• Differential expression is limited to basic statistical tests between manually selected groups

BioTuring BBrowserX + SpatialX

Overview: BioTuring BBrowserX + SpatialX is a commercial cloud platform for single-cell and spatial transcriptomics analysis. BioTuring positions itself as an enterprise-grade solution with GPU-accelerated analysis features. The platform combines data exploration with computational analysis tools and places emphasis on speed and ease of use.

Data import and format support: BBrowserX supports common single-cell formats, including .h5ad, Seurat objects, 10x HDF5, and MTX bundles. For spatial transcriptomics, it supports data from Visium, Xenium, Vizgen MERFISH, NanoString GeoMx DSP, and Slide-seq. The software also supports CITE-seq/ADT multi-omics data and TCR/BCR immune profiling.

Processing pipeline: BioTuring provides an end-to-end analysis pipeline that covers QC, normalization, dimensionality reduction, clustering, and downstream analysis. The platform emphasizes speed, leveraging GPU acceleration and optimized algorithms for fast processing of large datasets. Batch correction is available (Harmony).

Cell type annotation: BioTuring maintains its own proprietary cell-type database and offers automated annotation built on a curated database of over 100 million single-cell profiles, covering 54 cell types and 183 subtypes. Users can compare their data against public reference datasets to transfer labels.

Downstream analysis: BBrowserX includes differential expression analysis (with a proprietary method called Venice, alongside standard tests like t-test, Wilcoxon), gene set enrichment (via the fgsea package), trajectory analysis (via Monocle 2), and cell–cell communication analysis leveraging the CellChat ligand–receptor database. For spatial data, SpatialX adds neighborhood analysis, region-based segmentation, and spatially-aware differential expression.

Visualization: BioTuring provides interactive UMAP/t-SNE plots, spatial tissue views, heatmaps, violin plots, and dot plots. For figure customization, BioTuring offers a companion tool called BioVinci that supports themes and styling.

Data privacy and deployment: The cloud-hosted solution enables sharing and collaboration but involves uploading data to BioTuring's infrastructure. An enterprise on-premises deployment option is also available for institutions with strict data governance requirements.

Pricing: BBrowserX offers a limited free tier with basic features. Full analytical capabilities require a paid BBrowserX Pro license, which is typically priced on a per-seat basis, while enterprise pricing is available for institutional deployments. Overall cost can be higher, as the infrastructure requires GPU resources.

Pros

• End-to-end analysis pipeline with GPU acceleration for fast processing
• Supports multiple omics modalities: scRNA-seq, CITE-seq, spatial, TCR/BCR
• Wide format support (h5ad, rds, HDF5, MTX, CSV, spatial files)
• Proprietary cell-type prediction with public reference
• Enables institutional collaboration with data governance controls

Cons

• Pricing can be expensive as GPUs are required
• Cloud functionality involves uploading data to BioTuring servers, which may conflict with institutional data policies
• Not open-source

Pricing Comparison

Cost is often a decisive factor, especially when software budgets are limited. Here is a side-by-side breakdown of what each tool costs and what you get.

CellxGene: Free

CellxGene is entirely free and open-source under a permissive license. The hosted CZ CELLxGENE Discover portal and the local Explorer are all available at no cost. There are no tiers, no seat limits, and no feature gates. This makes CellxGene the most accessible option from a budget perspective, though it is limited to data exploration (not end-to-end analysis).

10x Genomics Loupe Browser: Free

Loupe Browser is free to download and use with no licensing fees. The only practical restriction is that it works exclusively with 10x Genomics data (.cloupe files generated by Cell Ranger, Space Ranger, etc.). For labs already on the 10x platform, this represents significant value: a polished visualization tool at zero marginal cost.

OmnibusX: Transparent published pricing

OmnibusX publishes its pricing publicly. All plans include a free 2-month trial with full feature access.

OmnibusX also offers a bioinformatics services add-on with a credit-based system (starting at $100 for 100 credits, with volume discounts up to $900 for 1,200 credits) for users who want hands-on analytical assistance from the OmnibusX team. All plans include unlimited scheduled support calls, which is notable — many comparable tools charge extra for direct access to bioinformatics expertise.

BioTuring: Quote-based pricing

BioTuring does not publicly list prices. Licensing requires contacting sales for a custom quote. The lack of published pricing can be a friction point for academic labs that need to plan budgets or compare costs across vendors before engaging with sales teams.

Summary

Each of these tools fills a distinct niche in the single-cell and spatial analysis landscape. The right approach depends on your data, your team's computational expertise, your institutional constraints, and your budget.

This comparison is based on publicly available documentation, published papers, and product information as of early 2026. Features and pricing may change. We encourage readers to evaluate trial versions and consult vendor documentation for the most current information.