TileDB's performant, scalable and cloud-native solution is ideal for our efforts at CZI to assemble and distribute one of the largest repositories of single-cell biological data in the world
Store data once as TileDB arrays, and enjoy interoperability with popular tools. We partnered with the Chan Zuckerberg Initiative to develop a data model that brings together the single-cell analysis toolchain.
Parallelize complex workloads and create distributed algorithms using TileDB Cloud. Access policies allow your work to stay within your organization, or securely share it with external collaborators to unlock atlas-scale analysis.
No more huge downloads and loading large datasets into memory. Cloud-native TileDB arrays allow you to slice straight from remote storage. Reduce costs and processing time by utilizing cost-efficient object storage services like S3.
TileDB arrays are a natural fit for the highly sparse data generated by single-cell platforms. Manage RNA-seq data and count matrices in a cloud-native format, and store any number of samples in a compressed and lossless manner for tremendous storage savings.
Through our partnership with the Chan Zuckerberg Initiative on the Unified Single-Cell Data Model and API, TileDB implements a common, open data model. Use Seurat, Bioconductor and Scanpy, integrated directly with TileDB data.
Eliminate data wrangling and large downloads. Slice TileDB arrays directly using native query conditions on genes and cells in any of TileDB’s language APIs, including R, Python and C++.
Preserve spatial context. In addition to sparse data, the TileDB array format excels at representing dense data types like biomedical imaging. As your research grows, combine datasets, code, ML models, files and more into TileDB Cloud groups to manage large projects.
Work across toolchains and languages to rapidly iterate using TileDB Cloud’s hosted Jupyter environment. For large computations, serverless user-defined functions and task graphs in TileDB Cloud let you parallelize custom algorithms efficiently and cost-effectively.
Create your own repository of single-cell datasets using TileDB Cloud. Make cloud datasets discoverable with cataloging features that help attract new collaborators. Securely share data in private, or make it public in a predictable model that eliminates surprise infrastructure bills.
We partnered with CZI to deliver the spec of an open source data model and corresponding implementations in R and Python for single-cell biology.
SOMA is a flexible, extensible, and open-source data model for representing annotated matrices, commonly used in single cell biology.
TileDB-based implementations of the SOMA data model that allows you to slice data directly from the cloud into Seurat or Bioconductor in R, and AnnData in Python.