Biology Reference
In-Depth Information
swissPIT project are aggregated, controlled, and coordinated by Grid
infrastructure, which is based on the ARC middleware.
53
It provides user-
level interfaces and abstractions of a seamless, homogeneous distributed
system. A high-level Web portal to swissPIT manages the Grid security
certificates on behalf of the users, thus sparing the scientists from having
to deal with credential delegation and renewal.
4. Proteomics Knowledge Integration
and Databases
A proteomic study typically generates a huge list of identified, and some-
times relatively quantified, proteins. However, a list of hundreds, or even
thousands, of independent items is actually poorly informative when one
aims at selecting a few potentially relevant biomarkers for further validation
studies. At this stage, a considerable amount of work still has to be done
on two fronts. On the one hand, database search mechanisms need
improvement to increase the knowledge on individual proteins and to
help sort out the candidates with the highest value
30
; these issues have
been described in the previous section. On the other hand, the repre-
sentation of data and knowledge requires further reflection to create
more meaning, as it enriches the interpretation of experimental results.
54
Characterization of the selected proteins is the next stage and is often
a very difficult task. It involves not only determining the precise form
of each protein (e.g. splice variant, phenotype, posttranslational pro-
cessing), but also investigating their possible interactions. The latter
studies raise a number of issues common to many bioinformatics appli-
cations, given that sequences constitute the core data type for all of
these applications.
4.1. Standards for High-Throughput Data
High-throughput analytical methods have produced an ever-growing
flow of new data. This situation has spawned a number of initiatives to
address the issue of data storage and data standardization. The transcrip-
tomics community has led by example, introducing a data format termed
MIAME (Minimum Information About a Microarray Experiment).
