blastnLABSCO SUMMARY
BLAST from the conversation, against databases you keep.
Point it at a local BLAST+ install and ask in plain language: search a sequence or a FASTA file against nt, nr or a database you built, and get the hits back in the format you need. For searches that take minutes, a queue mode hands back a job ID instead, but it needs a separate queue service.
What it is
A Python MCP server that drives a local NCBI BLAST+ installation. It runs nucleotide and protein searches, builds databases from your own FASTA files, and returns results in JSON, XML, tabular or pairwise form. Job-based searches for long runs need its separate queue mode.
What you get
- A nucleotide search (blastn) against a named database or a path
- A protein search (blastp) the same way
- A BLAST database built from your own FASTA file (makeblastdb)
- Job-based versions of both searches, with status checks and result pickup, in queue mode only, which needs the separate bio-mcp-queue service running
- Input given as either a FASTA file path or a raw sequence
- Output format chosen per search — tabular, XML, JSON or pairwise — with a limit on how many hits return
Requirements
BLAST+ installed on the machine — conda install -c bioconda blast, brew install blast, or apt-get install ncbi-blast+ — plus the server itself: clone the repository and run pip install -e . inside it. The client configuration starts local mode; job-based searches need the separate bio-mcp-queue setup and python -m src.main --mode queue. Databases are local: give a name such as nt or nr, or a path.
Setup effort
Follow the published steps — conda install -c bioconda blast
USE CASES
01
Identifying an unknown nucleotide sequence
Paste the sequence or give a FASTA path, search it against nt, and read the top hits with their E-values.
02
Finding homologues of a protein
Run a protein search against a database such as SwissProt and get the similar sequences back with alignments.
03
Searching against sequences you keep locally
Build a BLAST database from your own FASTA file, then search against it and read the hits in the format your pipeline wants.
7 tools
Listed in project docs. Your client may show a different list.
blastpmakeblastdbblastn_asyncblastp_asyncget_job_statusget_job_resultBEFORE YOU INSTALL
What to know first
Tools you get
7 tools named in the project's documentation.
Requires
- Python 3.9 or later on PATH
- BLAST+ installed, so blastn is on PATH (conda install -c bioconda blast, brew install blast, or sudo apt-get install ncbi-blast+)
- Local BLAST databases to search, such as nt or nr (BLASTDB points BLAST+ at them)
Where it runs
The documented client runs the server on your machine.
INSTALLATION
Connect your client
Server setup
Install BLAST+ (or with Homebrew: brew install blast; on Ubuntu: sudo apt-get install ncbi-blast+):
TERMINAL
conda install -c bioconda blastClone the repository and install the server:
TERMINAL
git clone https://github.com/bio-mcp/bio-mcp-blast.git
cd bio-mcp-blast
pip install -e .Add the configuration below to your MCP client; <YOUR_PATH_TO_BIO_MCP_BLAST> is the absolute path of the folder you cloned.
MCP Clients
Any MCP client
ANY MCP CLIENT · JSON
{
"mcpServers": {
"bio-blast": {
"command": "python",
"args": [
"-m",
"src.server"
],
"cwd": "<YOUR_PATH_TO_BIO_MCP_BLAST>",
"env": {
"PYTHONPATH": "<YOUR_PATH_TO_BIO_MCP_BLAST>"
}
}
}
}
