🚧 Work in Progress: This project is currently in the Alpha stage. Features, APIs, and performance characteristics are subject to change.
LOSAT (LOcal Sequence Alignment Tool) aims to deliver a lightweight, high-performance reimplementation of the NCBI BLAST algorithm, written in Rust and optimized for WebAssembly (Wasm) environments.
While originally designed to power web-based bioinformatics tools, LOSAT functions as both a standalone command-line interface (CLI) and an embeddable library. It aims to democratize sequence alignment by enabling it to run anywhere without external dependencies.
- Client-Side Processing: All computations happen locally on the user's machine.
- Data Sovereignty: Sensitive genomic data never needs to be uploaded to third-party servers, ensuring compliance with strict data privacy regulations.
- Serverless Architecture: Removes the need to maintain expensive backend infrastructure for alignment tasks.
- Portability: Compiled to WebAssembly, LOSAT runs natively in any modern web browser, regardless of the underlying operating system.
- Low Latency: Eliminates the network overhead (uploading/downloading large FASTA files) associated with traditional server-side BLAST.
- Embeddable: Designed to be easily integrated into Rust applications or JavaScript/TypeScript web frameworks.
Currently, LOSAT focuses on nucleotide-nucleotide comparisons:
- BLASTN: Standard nucleotide vs. nucleotide alignment.
- TBLASTX: Translated nucleotide vs. translated nucleotide alignment (useful for finding distant relationships in coding regions).
Protein alignment modes (BLASTP, BLASTX) are planned for future updates.
Integrating sequence alignment into web applications has traditionally been a challenge. The industry-standard NCBI BLAST+ is a command-line tool dependent on the host OS, making it impossible to run directly within a browser sandbox.
This limitation forced developers of tools like gbdraw (a genome visualization platform) to impose a fragmented workflow on their users:
- Export sequences from the web app.
- Process them externally using CLI tools or third-party web services (e.g., NCBI Web BLAST).
- Import the results back into the application.
LOSAT breaks this cycle. By porting the core alignment logic to Rust, we can compile the algorithm to WebAssembly. This allows gbdraw and similar tools to perform alignments internally and instantly. Users perceive the analysis as a native feature of the application, dramatically improving the user experience.
LOSAT is not a wrapper around existing C++ code but a native Rust implementation.
- Memory Safety: Leverages Rust's ownership model to ensure memory safety without a garbage collector.
- Wasm-Bindgen: Uses
wasm-bindgento facilitate high-speed communication between JavaScript and Rust modules.
LOSAT is optimized for "local" alignments (e.g., comparing two viral genomes or gene clusters) rather than querying massive databases like GenBank/nt.
- Database Size: Best suited for small-to-medium datasets loaded in memory.
- Features: Advanced BLAST+ parameters (like masking algorithms or specific output formats) may not yet be fully implemented.