No description, website, or topics provided.
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
SVGen @ c4e6a4b
Simulation
TwoPaCo @ 137d8df
gfatools @ 7693c72
gfautils
grc
html
k8 @ b646aa3
sequenceTubeMap @ 556bc6c
.gitmodules
LICENSE
Makefile
README.md

README.md

Graphy McBrowserface

This is a toy example of a graph genome browser.

Motivation

To date, the reference genome assemblies for most species, including human, have been represented as linear, haploid sequences. With the advent of population-scale sequencing, graph structures have been proposed as a better way to represent genomes. A graph has the advantage of naturally representing variation as forking edges (“bubbles”) and individuals as paths through the graph. Read alignment to graph genomes, and variant calling from those alignments, has been shown to be both more sensitive and more specific than using linear genomes. However, the migration to graph-based genomes is hindered by 1) the non-interoperability of most existing bioinformatics tools with sequence graphs; 2) the substantial challenge and degree of effort involved in adapting those tools; 3) new challenges in annotating and visualizing graph-based references and alignments. Genome browsers are among the most widely used tools among both novices and experts in genomics. To motivate the transition to graph genomes, we developed a simple graph-based genome browser that is accessible at http://graphymcbrowserface.umbc.in.

Results

We did the following:

  1. Generate a population of chromosomes (GRCh38 chr22) with randomly-generated structural variants inserted, using SVGen.
  2. Build a population graph of the generated chromosomes plus the reference genome, using TwoPaCo. Dump the graph to GFA.
  3. Create a GFA to JSONGraph converter.
  4. Implement several different Javascript-based viewers for JSONGraph data, using:

Dependencies

  • JSON.pm
  • python (2.x)
  • Web Server (Apache, nginx etc) for web components

Installation

Use the Makefile to install SVGen and TwoPaCo.

export SRV_INSTALL_PATH=/var/www && make build && sudo -E make install

Setup

  1. Generate sequences and assemble the graph
# Generate a graph from 12 simulated sequences
export $SVGEN_BASE
svgen_generate_sequences $SVGEN_BASE 12
svgen_assemble $SVGEN_BASE
  1. Start the server
  2. Browse to XXX

Manuscript

Draft

Contributors