Biomedical Engineering Reference
In-Depth Information
Consortium with great mechanisms for storing microscope and other
technical meta-data. Omero provides excellent capability for connectivity
and there is a wealth of client applications that are able to interface with
it. With these applications a user can select images for manipulation,
annotation and analysis. Image analysis procedures such as segmentation,
object identifi cation and quantifi cation can be performed within the
client applications direct on the user's desktop machine pulling data
automatically from the Omero database.
11.6 Analysing the data: helping users
work with their own data
Once we have our data under control and in our apps/databases, we
visualise with external tools. For the majority of our next-generation
sequencing experiments, genome browsers are very popular with our
users. We completely eschew the use of GBrowse-style browsers [16] that
require a lot of setup on an external server and require the bioinformatician
to upload (and prepare) data for a user before it can even be looked at.
Instead we opt for the very useful and user-friendly SAVANT [17] and
Integrated Genome Viewer (IGV) [18] desktop genome browsers.
SAVANT works well in most cases, presenting a clean and straightforward
interface for viewing the alignment of next-generation reads on reference
or custom assemblies, which is mostly what we want. More can be done
with IGV, it is particularly suited to fi ltering and comparing data. Both
work well but have some minor network issues, getting either to start
from behind a fi rewall or proxy is initially diffi cult. On fi rst loading in
such an environment SAVANT takes a while to initialise (long enough to
make a user think the machine has crashed), patience and turning off
auto-update once it is loaded will make subsequent loads much quicker.
IGV relies more on fi les downloaded over the internet and needs details
of the proxy to function. Early versions of SAVANT could crash with an
out-of-memory error when trying to load in reference sequence, later
versions have fi xed this problem.
For some applications, such as hosting a next-generation data set to be
viewed by collaborators over the web, we have used other web service-
style browsers like JBrowse [19] (part of the GMOD project) and AnnoJ.
Both of these are fast, fl exible and somewhat attractive, but neither is
anywhere near as easy to setup as the desktop browsers above. AnnoJ
[20] is a web service that renders feature data. It makes requests of another
Search WWH ::
Custom Search