tabix-service

A nodejs-based service for requests for tabix-indexed genomic data

Prerequisites

These instructions assume a CentOS 7 host and sudo-level access. It's probably easiest to set up a virtual host and run these tests within that host.

htslib toolkit

To set up htslib (and bgzip and tabix tools):

$ git clone git://github.com/samtools/htslib.git
$ git clone git://github.com/samtools/bcftools.git
$ sudo yum install zlib-devel
$ sudo yum install bzip2-devel
$ sudo yum install xz-devel
$ sudo yum install curl-devel
$ cd bcftools
$ make
$ sudo make install
$ cd ../htslib
$ make
$ sudo make install

This should put bgzip and tabix into /usr/local/bin.

Compressing and indexing test intervals

Compress and index the test interval file sample.bed, if not already done:

$ bgzip sample.bed
$ tabix -p bed sample.bed.gz

This creates two files: sample.bed.gz and sample.bed.gz.tbi.

A test query would work like this:

$ tabix sample.bed.gz chr1:10000000-11000000
chr1	10037856      10038006	id-501897	102
chr1	10197056      10197206	id-509857	19 

Node.js

To install Node.js:

$ sudo yum install epel-release
$ sudo yum install nodejs
$ sudo yum install npm

Tabix server and client

Install the Node.js dependencies, if not installed:

$ cd tabix-service/server
$ npm install
$ cd ../client
$ npm install

Starting Tabix service

$ cd tabix-service/server
$ export TABIX_SERVER_PORT=1234
$ node app.js

Starting Tabix test client

$ cd tabix-service/client
$ export TABIX_CLIENT_PORT=4567
$ node app.js

Then visit http://localhost:4567 to run test queries.