Exome browser

PLINK/Seq has a simple, table-based browser called pbrowse, designed for viewing and searching raw project data. It can either be run locally (where the data are expected also to reside locally), or the same code can be hosted on a webserver and run remotely via CGI (where the data are expected to reside remotely, on a filesystem accessible from the webserver).

pbrowse instantiation of 1000 Genomes phase 1 data (16-March-2012 release)

An example project is hosted here: the subset of all RefSeq coding variants from the 1000 Genomes phase 1 dataset (16-March-2012 release) -- i.e. an extracted "whole-exome" dataset from this whole-genome study.

  • Read the instructions on the first page
  • Enter GeneID CES1 to pull up variants in that gene (note: only a minority of all genes are represented in the pilot 3 data)
  • Click on view to see individual genotypes for one of the variants
  • Enter phe1 in the phenotype box to append the "phenotype" (CEU/TSI) and "case/control" counts
  • Enter * in the meta-field to append all meta-tags
  • Experiment with the mask field, e.g. mac=1-2 to view only variants seen once or twice
  • Clear the GeneID box and enter chr1 in the rightmost regions box, for example, to view all vaiants on chromosome 1

Running the browser locally for your own projects

The browser can either be run locally, or sitting on a web-server. After downloading the binary for PSEQ, the browser should be included. Assuming you've copied the files to somewhere in your command path as instructed, you run the browser:

pbrowse /path/to/my/project

This starts up a lightweight http server (mongoose) and uses it to run the CGI-based pbrowse.cgi. The script opens up firefox as the default web-browser. You can easily change this by editing the small pbrowse file (it is just a bash script).

More flexible tools for interactively working with project data are under development.