The beginning of what looks to be an interesting series of articles on hacking the human genome has been posted on O'Reilly's MacDev Center. Some interesting quotes:
On the UC Santa Cruz Genome Browser: "I'm not sure what Edward Tufte would have to say about the design of this, but there are a lot of annotations available and you need to see all of them to evaluate any given region."
There are a lot of well-characterized bacterial genomes available, and comparison of these gets really interesting.
I've been told that comparison of archaeal genomes is also "really interesting".
Comments
It only makes sense comparatively
My personal favourite quote: "Another might be that a royal family has no function whatsoever in a modern society and survives in London merely as an evolutionary remnant".
Is this to be a series? Hope so, but it's not clear. Also by the authors admission, there's not much specific to OSX here. I think a rather better article is this one from Linux World, detailing how Linux and open source benefitted the sequencing and analysis of the SARS virus.
Having spent the past month comparing my incomplete, draft genome to about 170 others has certainly opened my eyes to many things. The first is that visualisation can certainly be a useful exercise. The second is the sheer number of custom scripts I had to write to make this happen. Not so bad for completed genomes (thanks to Bio::DB::GFF and Bio::Graphics), but someone really needs to come up with a schema for draft genome data.
I'm glad someone else sees fit to criticise the UCSC browser. It is a nice piece of work with some good features, notably the ability to incorporate BLAT tracks...but it's pretty hard to see anything of use with it in my limited experience, because there's too much stuff per track and it's too small. I much prefer the Generic Genome Browser - free, open source, integrated with the Bioperl modules I mentioned earlier, plenty of features and nice on the eye. Try some microbial genomes at my server.
At Your Server?
Not Found
The requested URL /cgi-bin/gbrowse was not found on this server.
Apache/1.3.33 Server at psychro.bioinformatics.unsw.edu.au Port 80
moved
Just found this comment. We moved to mod_perl, the new link is here.
Visualization of views gets complex
I'm really glad that someone else echoes my opinion of the UCSC browser. Comparative genomics with qualitative data requires an extensible viewer that meshes well with bioinformatics sequence processing modules. I've been using the Generic Genome Browser for some time and am in the process of migrating to the GMOD schema, which is working quite nicely. However, getting to this point (MSAs, microarray data, nonparametric scoring) was a real pain, script-wise.
While the GGB has the nice feature of allowing on-the-fly processing of genomic element data according to the "predictive scoring method of the day" through CGI, it still falls prey to the same limitations of the UCSC browser in that it is web-based, and while zoomable, not scrollable. Considerable work has to be done to allow dynamic view links to the underlying database to be constructed depending upon what the user (viewer) is looking at at any given moment. Since I work in a non-coding RNA prediction environment, the natural language processing for "isa" or "hasa" is troublesome (but thankfully somewhat handled by the GFF format). A scrollable ontological map viewer may be more amenable to the transition from genomic element position to scoring methods and natural language ontologies. Better yet would be a viewer that did not have to refresh, but still loaded memory in manageable chunks at a time for rapid initialization.
-my2cents
The major difference between a thing that might go wrong and a thing that cannot possibly go wrong is that when a thing that cannot possibly go wrong goes wrong, it usually turns out to be impossible to get at or repair.
genome browser software
If dynamic and flexible whole-genome viewing is important to your project, I recommend taking a look at the Integrated Genome Browser, an open source viewer released by Affymetrix.
The URL for the project is:www.genoviz.org.
A Java Web Start version is available here:
www.affymetrix.com/support/developer/tools/download_igb.affx
I use this viewer for in-depth analysis of my own data sets. I load in my data and then use the "Load DAS" feature to populate the scene with data from the UCSC Distributed Annotation Server, which serves the same annotations available in the UCSC Web-based viewer. This lets me view my own annotations together with the public data, an essential feature if you are developing new annotation methods and need to be able to scan quickly through them. The Santa Cruz browser is a little too clunky for doing genome-scale viewing.