Biomedical Engineering Reference
In-Depth Information
Installation, operation, and support of a local Atlas installation has
proven to be very feasible and is certainly a cost-effective option to
consider. Supplementing internal data has also proven to be feasible and
worthwhile. As will be discussed in the next section, efforts to reduce the
amount of manual curation in data entry are under active development.
9.4.4 Next steps and future considerations for
an in-house Atlas
Manual curation of experimental meta-data for custom studies has
proven to be feasible, but this is certainly not a high-value activity for
in-house resource and is prone to human error. It would be highly
desirable to map standard terms used for annotation of in-house data
with the EBI EFO. Better yet, the EFO could be used in systems to capture
experiment and sample meta-data so that data are complete and captured
in a standardized format from the start! As the EFO is inherently dynamic,
update and exchange mechanisms would be highly desirable to bring in
new terms of general interest such as standard strain, cell line and
treatment names.
As described above, public studies of interest to Pfi zer have been
curated in-house and loaded into the local Atlas. Although this activity
adds value, it is likely to be of pre-competitive value to Pfi zer and other
commercial entities. Therefore, it would be of mutual benefi t to public
and private institutions having local installations to be able to publish
locally curated public studies back to the public Atlas, subject to review
of curators of the public Atlas. As a considerable portion of the available
public expression data is currently uncurated, it would be mutually
benefi cial for the Atlas community to consider a one-time effort to review
and curate all public studies meeting a defi ned quality standard. Such a
project would benefi t commercial and non-commercial partners alike,
and it would be fi tting to have an open source software system developed
to share public expression data contribute to an open science model for
gene expression and other 'omics data.
￿ ￿ ￿ ￿ ￿
9.5 Conclusion and learnings
The open source EBI Atlas platform has proven to be a great project to
bring curated expression studies into an easy to use portal system for
scientists. Initial analysis of a gap in vendor solutions in this area pointed
 
Search WWH ::




Custom Search