Researchers Released Gravlax Tool for RNA Data
A new computational method enables faster, smaller archiving and re-quantification of single-cell RNA sequencing data.
Updated on Sept. 21, 2026 in Quantum Computing
Live Poll
Do you believe new archiving tools make scientific research more reliable for the general public?
Scientists have introduced Gravlax, a computational tool designed to archive single-cell RNA-seq molecular data efficiently. This method allows users to re-quantify data under updated gene annotations without the need to re-process original raw files.
Why it matters
Current count matrices are tied to specific gene annotations, making it difficult to reinterpret data as new biological knowledge emerges. Gravlax solves this by deferring annotation-dependent decisions until the analysis phase.
Gravlax achieves storage of 11 to 18 bits per read, resulting in archives 9.0 to 12.7 times smaller than tag-preserving CRAM. Count matrices replayed via the tool deviate from direct STARsolo quantification by only 0.24 to 0.75 percent.
The players
Gravlax
This is a computational tool developed for archiving single-cell RNA-seq molecular evidence using a compact, content-authenticated format.
COMBINE Lab
This research institution manages the source code repository for the Gravlax tool.
The details
The tool works by storing essential molecular relations such as shared genomic geometry, barcode identity, and UMI equality. It is implemented in Rust and enables researchers to identify unannotated splice events across donor cohorts in just 9 seconds.
Timeline
September 21, 2026: The research describing the tool was published.
The Tech Race
The introduction of Gravlax represents a shift toward more flexible genomic archiving that moves away from static, annotation-locked count matrices. This approach accelerates large-scale data re-analysis, positioning it as a competitor to legacy processing pipelines like STARsolo.
Bioinformaticians and researchers can now store significantly larger datasets while reducing the time required to re-quantify data by up to 82 times. This efficiency allows for faster discovery of novel splice events and easier integration of updated gene references.
The takeaway
Gravlax provides a more sustainable model for genomic data storage by decoupling raw molecular data from fixed annotation versions. This innovation suggests that future biological archives will prioritize re-interpretable formats to keep pace with rapid genomic discoveries.
Further reading
Learn more about the latest innovations in high-performance genomic processing in our Quantum Computing section.
More information
Access the technical documentation and code for this method at the gravlax source code repository.
Live Poll
Do you believe new archiving tools make scientific research more reliable for the general public?
