News

TileDB x Databricks Partner to Power Multimodal Data for Agentic AI in Healthcare + Life Sciences. Read the news

3 min read

Data Management
Data Science
Single Cell

Single-cell analysis software designed for discovery

Originally published: Oct 16, 2025

Table Of Contents:

Powerful single-cell analysis tools built to scale

Single-cell RNA-seq analysis software to unify your data

What sets TileDB single-cell sequencing analysis software apart

Maximize the potential of your transcriptomics data with TileDB

Start unifying your single-cell data in one platform

To unlock the potential of single-cell RNA sequencing, you need scalable, efficient, and easy to use single-cell analysis software built for the complexities of transcriptomics data. TileDB unifies the different formats of single-cell data with a common data model at atlas scale, which  the Chan Zuckerberg CellxGene Atlas trusts to power FAIR data from their 100M+ sequenced cells. The result is a single-cell analysis platform that empowers researchers to focus on scientific discovery rather than tedious data management.

Simplify your single-cell analysis

Powerful single-cell analysis tools built to scale

TileDB Carrara is a powerful and flexible database solution architected on multi-dimensional arrays, and it powers TileDB SOMA, a specialized database system designed for managing single cell data. TileDB SOMA (Stack of Matrices, Annotated) makes it easy for your  researchers to use their favorite single-cell analysis tools inside a scalable, efficient, and user-friendly solution built for atlas-scale data.

Interoperable with preferred single-cell analysis tools

TileDB-SOMA works seamlessly with both Python and R, and tightly integrates with Seurat, Bioconductor and Scanpy so single-cell data scientists can work in the languages they prefer.

Optimized for cloud object stores

TileDB-SOMA’s cloud-native array format is specially optimized for object stores like Amazon S3, Google Cloud Storage, Azure Blob Storage and MinIO. This helps drive efficiency on your choice of public cloud infrastructure for simplicity and cost savings.

Designed for single-cell scale 

TileDB-SOMA proves its scalability by handling data from 100M+ cells in the CZ CellxGene Atlas, easily managing the large data quantities required by single-cell research through TileDB Carrara’s distributed computing engine.

Vector search for cell similarity and annotation

TileDB's vector search capabilities enable automated cell type annotation and interactive analysis for deeper biological insights, helping single-cell researchers streamline reference mapping workflows to explore data from expressed genes more efficiently.

Single-cell RNA-seq analysis software to unify your data

TileDB Carrara is designed for the complexities of RNA-seq data analysis. It effortlessly unifies single-cell data from all formats and tools inside one secure platform to easily store, browse, share, and query transcriptomic data from differentially expressed genes.

How TileDB single-cell analysis empowers research teams:

  • Transition seamlessly from file storage to active analysis TileDB Carrara makes it easy for bioinformaticians to load single-cell output files directly into a notebook and convert them into structured TileDB-SOMA assets. Besides simplifying data harmonization, this optimizes storage costs and query performance.

  • Assemble an efficient FAIR data catalog Therapeutics company Cellarity chose TileDB as their FAIR platform to support their deep learning models by analyzing transcriptomic data from hundreds of millions of single cells. This helped Cellarity build in less than an hour a single-cell atlas that used to require days.

  • Train foundation models with single-cell data With the TILEDB-SOMA-ML library, your researchers can use PyTorch to train AI/ML models at scale, with optimized data workflows for efficient shuffling, sampling, and analysis of scRNA seq datasets and other complex data.

What sets TileDB single-cell sequencing analysis software apart

TileDB Carrara empowers single-cell sequencing analysis software through infrastructure built to master complex multimodal data and accelerate research discovery. Here’s how TileDB Carrara stands out:

  • Reference files become easily queryable databases: Carrara maintains files’ familiar folder hierarchies and operations, but adds capabilities like searchable metadata, programmatic API access, secure sharing, and governance. In addition, native support for Nextflow workflows enable users to go directly from raw sequence data (like scRNA seq) to queryable databases. By creating logical mappings instead of copying data, Carrara makes it easy to run complex queries while keeping storage costs to a minimum.

  • Teamspaces for simple, secure collaboration: TileDB Carrara’s Teamspaces set up project-level environments enabling teams to manage permissions and security for data access. After files are mapped into a teamspace, Carrara handles all the cloud security configuration, including identity and access management, audit logging, and compliance tracking—all without moving or copying your data.

  • Support for spatial transcriptomics: TileDB Carrara provides built-in support for spatial transcriptomics to combine the precision of single cell RNA data with the context of imaging data for determining spatial relationships. This enables researchers to efficiently write and access large spatial datasets both locally and in cloud storage through a centralized data store optimized for long-term cost efficiency.

  • High-performance table operations: Just as TileDB excels at handling complex data structures, Carrara also delivers exceptional performance in traditional table operations. Analysts can write SQL directly against TileDB-SOMA tables to run aggregations and transformations on observation (cell) and variable (gene) matrices. Through Carrara’s array-native operations, you can leverage a holistic analytical environment inside a unified workspace.

Maximize the potential of your transcriptomics data with TileDB

Because TileDB’s technology is architected around multi-dimensional arrays, it acts as a universal database platform for transcriptomics data like bulk RNA-seq as well as gene expression data, imaging, EMRs and other data modalities. By managing all your multimodal data inside a single system, TileDB Carrara helps you eliminate silos, improve processing performance, reduce storage costs and, most importantly, speed analysis that leads to discovery.

Start unifying your single-cell data in one platform

Ready to unify your single-cell data inside one management platform? Contact us to see how TileDB Carrara can empower your transcriptomics data management.

Meet the authors