nmdp-bioinformatics / imgt_biosqldb

The IMGT/HLA sequence data persisted in a BioSQL database

imgt_biosqldb

The IMGT/HLA sequence database persisted in a mysql BioSQL DB

Overview

This repository contains code for building a BioSQL database with the IMGT/HLA sequence data. The database is generated with the create_imgtdb.py script and is built in a Docker image using the Dockerfile. This database is utilized by the sequence annotation (SeqAnn) python package for quickly retrieving sequence records as BioPython SeqRecord objects. Please refer to the documentation on the IMGT/HLA GitHub repository for any questions about the sequence data.

Docker

This database can be run locally by building it from the Dockerfile or by pulling a prebuilt image from Docker Hub. Pulling from Docker Hub is the quickest option if you don't need to increase the number of release loaded into the database.

Pulling from Docker Hub

docker pull nmdpbioinformatics/imgt_biosqldb
docker run -d --name imgt_biosqldb -p 3306:3306 \
    -e MYSQL_ROOT_PASSWORD=my-secret-pw nmdpbioinformatics/imgt_biosqldb

The -d flag runs the service in "detached-mode" in the background and -p specifies what ports to expose. Make sure the ports you expose are not already in use. If the docker container is successfuly executed then typing docker ps -a will show a new container labeled imgt_biosqldb running. In order for this to run you must set the MYSQL_ROOT_PASSWORD variable. Click here for more information on the publically available docker image.

Building locally

git clone https://github.com/nmdp-bioinformatics/imgt_biosqldb
cd nmdp-bioinformatics/imgt_biosqldb
docker build -t imgt_biosqldb:latest --build-arg RELEASES="3310,3320" .
docker run -d --name imgt_biosqldb -p 3306:3306 \
    -e MYSQL_ROOT_PASSWORD=my-secret-pw imgt_biosqldb:latest

This will build the docker image locally but changes the number of IMGT/HLA RELEASES loaded from three to five. This has only been tested for the last five IMGT/HLA releases. Some of the formats start to change a little when you go father back with the IMGT/HLA releases. Therefore, errors may occur when trying to build with more than five releases.

Accessing the database

# pip install pymysql biopython
>>> import pymysql
>>> from BioSQL import BioSeqDatabase
>>> server = BioSeqDatabase.open_database(driver="pymysql", user="root",
...                                       passwd="my-secret-pw", host="localhost",
...                                       db="bioseqdb")
>>> db = server['3310_A']
>>> seqrecord = db.lookup(name="HLA-A*01:01:01:01")
>>> seqrecord
DBSeqRecord(seq=DBSeq('CAGGAGCAGAGGGGTCAGGGCGAAGTCCCAGGGCCCCAGGCGTGGCTCTCAGGG...AAA', DNAAlphabet()), id='HLA00001.1', name='HLA-A*01:01:01:01', description='HLA-A*01:01:01:01, Human MHC Class I sequence', dbxrefs=['EMBL:AJ278305', 'EMBL:AL645935', 'EMBL:CR759913', 'EMBL:EU445470', 'EMBL:GU812295', 'EMBL:HG794373', 'EMBL:M24043', 'EMBL:X55710', 'EMBL:Z93949'])

When you have the imgt_biosqldb docker container running locally you can access it with any programming language that has a mysql connector. This is an example of accessing the database with python using the BioSeqDatabase class from the biopython package. Please refer to the sequence annotation repository for more examples of how to use this database.

Related Links

• The Immuno Polymorphism Database (IPD)
• Sequence Annotation
• Feature Service
• GFE DB

IMGT Reference

James Robinson, Jason A. Halliwell, Hamish McWilliam, Rodrigo Lopez, Steven G. E. Marsh; IPD—the Immuno Polymorphism Database, Nucleic Acids Research, Volume 41, Issue D1, 1 January 2013, Pages D1234–D1240, https://doi.org/10.1093/nar/gks1140