Given some genomic data in a well-known format (e.g. GFF) with gene models, what tool(s) do you use to visualize that data.
What tools allow you to add your own tracks of data easily?
I'm interested in both desktop and web-based tools--with preference to those that are customizable via some kind of API.
I like the Broad's IGV: http://www.broadinstitute.org/igv/ for genome browsing. It handles lots of common data formats, including SAM/BAM files if you're dealing with NGS-scale datasets. Apparently you can talk to it it via http, but you can't say anything very complicated: http://www.broadinstitute.org/igv/PortCommands
A bit different from the linear browsers listed above is the genome visualization tool circos. It can plot a wide range of different data types onto the radially displayed chromosomes. Everything can be customized and is quite easy to use.
Late to the party here, but jbrowse has a nice ajaxy interface...
It's interesting to see votes for the IGV; I've looked at it only recently and it does look very promising as a desktop viewer.
I'm a long-time user of the Generic Genome Browser, a Bioperl-based web application. It is extremely customizable and can act as both DAS server and client. Be prepared to spend quite a bit of time "munging" your GFF files into shape and working on the config file to get the desired results.
I also like Bioperl's Bio::Graphics as a way to take simple text files (including GFF) and quickly generate very attractive plots. There's a similar, but less extensive Ruby library too.
I've recently tried GenomeGraphs, an R Bioconductor package. It fetches annotations from Ensembl and plots them as tracks. It's a good way to overlay quantitative data onto genomic features: here is a sample plot.
I'm using the well-known UCSC Genome Browser http://genome.ucsc.edu/goldenPath/help/customTrack.html
I also wrote an experimental implementation of Jan Aerts' LocusTree based on BerkeleyDB (see http://plindenbaum.blogspot.com/2009/11/java-implementation-of-jan-aerts.html)
Just to chip in my first comment on BioStar. I also use Gbrowse for this job. Hugely configurable and very easy to install and add tracks to. I've recently been using Gbrowse2, which is a pretty slick improvement. A comprehensive install guide is available at the GMOD website http://gmod.org/wiki/GBrowse_2.0_HOWTO
There's was a great review paper on genome visualization in Nature Methods last year: Visualizing genomes: techniques and challenges, by CB Nielsen, et al. It's part of a whole supplement on visualizing biological data.
And, for what it's worth, both GBrowse and JBrowse are both highly configurable and built to be installed locally, and have active communities supporting them.
Thanks for all the replies so far. I just want to add a couple I've found:
Annoj with a nice demo here: http://neomorph.salk.edu/epigenome/epigenome.html
seems to be a pretty nice web-interface and I've been able to get my own annotations drawn. it may be an abandoned project, and I wasn't able to find the javascript source (only the packed/minified version).
also genometools wraps cairo to allow simple drawing of regions: here's the image from their home page:
And they have bindings to a variety of scripting languages.
Anno-j has a really nice architecture. I used it to setup an Arabidopsis browser for custom experimental data along with their existing tracks: http://arab-annoj.appspot.com/. The downside is that I haven't seen any active development in a while, which is a shame.
I would vote for IGB (Integrated Genome Browser). It can handle a large number of tracks and large genome scale data sets. It is highly configurable and very fast. You can also talk to it via some ports . I sometimes generate excel files with hyperlinks that make IGB jump to specific locations, which can be really useful. So you can have an excel sheet with a list of genes which you can click in one window and IGB jumps to that gene.
Another great feature is the ability to do data manipulation, such as merging tracks using logic, to get overlapping or unique features. You can also do things like log2 transformations.
Another one I found today: GView: https://www.gview.ca/wiki/GView/WebHome
GView is a java package used to display and navigate bacterial genomes. GView is useful for producing high-quality genome maps for use in publications and websites, or used as a visualization tool in a sequence annotation pipeline. Users can interact with the genome using a powerful pan-and-zoom interface, or GView can write static images of a genome to a file. GView can draw a genome using either circular or linear layouts, with additional layout types planned for future release.
For desktop level genome visualization the Integrated Genome Browser is a very nice tool.
Scribl is based on HTML5 Canvas which allows for some nice interactivity (demo). Though it is a little light on features.
I'll put in a plug for Genboree, which my lab develops. It's essentially a genome browser with personalized databases/wikis, access control, and integration with Galaxy. It can import UCSC tracks and you can add any data that maps to genomic coordinates. It also has a REST API that's fairly new.
Since I do a lot of copy-number work, I also do quite a bit of sanity-check visualization with R, often just using simple tweaks of the plot command, but sometimes pulling in packages to help.
Circos would get my vote too.
Broad Institute's Integrative Genomics Viewer (IGV) will be my first choice, some other tools also to look at..
For visualization of genome re-arrangements, take a look at Gremlin. The authors compared it to Circos and found that Gremlin led to more and deeper insights for users than Circos.
Unfortunately it seems that this software is not yet available for download according to Ben Raphael's lab software page: http://compbio.cs.brown.edu/software.html
Dalliance is looking nice: http://www.biodalliance.org/
I am currently using LocusZoom for annotation of our GWAS data. It provides neat way to visualize crowded locus regions.
MagicViewer seems also to be a nice solution.
I like Argo from Broad Institute much more than IGV. It works well for visualizing and manually annotating whole genomes and it's reasonably fast. It can run standalone or as an applet/webstart (it's written in Java and distributed as a single jar file).
I'll throw another name into the ring - I've recently been using SeqMonk for some of this type of stuff: http://www.bioinformatics.bbsrc.ac.uk/projects/seqmonk/README.txt. I like its capabilities for doing quantitation over user-defined probes along your genome. There's lots of flexibility there for starting an exploratory enrichment-type analysis of your data.
Wow, there are lots of options to do this! One that hasn't been mentioned is the Ensembl Genome Browser. The positive would be the ability to compare your gene models to Ensembl annotation along the genome. For example, in this view.
Click 'Manage your data' at the left. Either upload a GFF file to Ensembl using the 'Upload data' option at the left of the menu, or use 'Attach URL' to attach your file without uploading it to Ensembl. In both cases, you will be able to view your gene models along the genome, and turn on or off other data tracks.
If you need any help with this option, give it a try or ask our helpdesk to guide you. helpdesk [at] ensembl [dot] org
One more from VizBi 2012: Chipster http://chipster.csc.fi/
One more (January 2013): ChromoZoom
ChromoZoom: a flexible, fluid, web-based genome browser Bioinformatics (2012) doi: 10.1093/bioinformatics/bts695
Also see the post about drawing chromosome ideograms
GenomeBrowse from Golden Helix is a desktop client.
Well, since this old thread has been revived I cannot help promoting my own tool.
ASCIIGenome is a genome browser based on command line interface and designed for console terminals.
There is one tool called Artemis which is developed by Sanger institute and one more tools called SeqVista
Hope this will help you.
One more via: https://twitter.com/klmr/status/227780333981028352
No love for Tablet? I'm quite surprised to not see it mentioned here. Of course, IGV is the classic, and Tablet is a bit simpler than that or other tools listed here, but it's a nice quick tool to visualize BAM alignments with GFF features. I use it often when I don't want to bother with IGV, where some stuff is just too complicated - I use IGV mostly for some advanced stuff.
in all fairness, Tablet did not exist when this thread was started. It is quite impressive that Tablet is still operational though, I gave it a go right now.
For reasons that I can't quite explain if find the output drawn by Tablet as surprisingly distracting, almost discomforting to look at.
It is "too pretty" - I feel gratuitously so. Too many little features drawn to each base, in its own little shaded little box etc. Yet I cannot easily tell which way the read aligns (forward or reverse) The reads have the wrong widths etc. I find it to impose a cognitive overhead.
Just a quick note on why Tablet might not be used more widely.
Interesting, I did not think about it like that. Tablet has several colour schemes in the 2nd tab (I think) to visualize different features of the alignment, including read orientation etc. These schemes actually let me focus exactly on what I need without any visual noise. Maybe that will help you too.
I also like the left sidebar, where I can quickly search contigs and even get basic info, like length or number of mapped reads, which is very usefull when having lower-quality data for non-model organism.
Another option is D3 Genome Browser (d3gb.usal.es). It supports the most popular genomic formats and you can easily add tracks with a single command in R or Python.
It is a multi-platform software for quick and in-browser visualization of genomic data and it is available as an R package, a Python module and a WordPress plugin.
Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
version 2.3.6
Thanks, I finally had a chance to look at this and while you can't say much, you can say enough to load some data, then (programmatically) go to a bunch of locations and save an image. very useful.