BIQ

Back-splice junction indexing and querying

About

BIQ is a tool for searching k-mers that span backsplice junctions (BSJs) in a large collection of RNA-Seq experiments. The main use case is to identify experiments in which a particular circular RNA, identified by its backsplice junction, is expressed. The method is described in the manuscript on bioRxiv.

BIQ builds on top of KIQ, a software for indexing and querying k-mers in sequencing data sets.

An example installation can be found at e-rna.org or at GitHub.
It contains three BSJ k-mer indexes:

1. Drosophila melanogaster, 103 RNA-Seq datasets from Westholm et al 2014.
2. Human, 386 SRA files from ENCODE (only total RNA-Seq).
3. Human, 2652 SRA files from Solomon et al. 2016.

Installation

BIQ is a React app built with create-react-app, which requires Node.js to be installed.

Download:

git clone https://github.com/pmenzel/biq
cd biq

First install the required packages:

npm install

Run the app locally in the development mode:

npm start

Open http://localhost:3000 to view it in the browser.

For preparing the installation on a web server, first run

npm run build

and then copy content of the build/ folder to the web server.

Configuration

BIQ accesses KIQ and the backsplice junction index through an HTTP endpoint. The folder http_endpoints contains HTTP wrappers for PHP and AWS Lambda that can be used to call KIQ. The URLs for these endpoints can be set in src/conf.js.

The enumeration of BSJ k-mers from annotated transcripts and their exons is done with the script scripts_ensembl/enumerate_BSJs.pl.

BIQ also needs a JSON file containing the 5' and 3' sequences of the exons, which are used to create a query k-mer. This JSON file is created with the script scripts_ensembl/exons2json.pl.

License

Copyright (c) 2018,2019 Peter Menzel pmenzel@gmail.com

See the file LICENSE.