Biology Reference
In-Depth Information
elucidate patterns of correlation between information terms and user-defined
input interrogator concepts within an unstructured collection of text. LSI relies
upon the potentially meaningful association of words/items within a body of
text. With the creation of a matrix of scientific (e.g., PubMed abstracts) text-
to-word associations, it is possible to identify and quantitate even noncurated/
experimentally derived links between input concepts and a specific gene or
protein.
141,144,145
The application of LSI-based techniques to biological sys-
tems will greatly assist the discovery process for signaling network analyses
and has recently been demonstrated to effectively synergize with existing
human-curated KEGG (Kyoto Encyclopedia of Genes and Genomes,
novel
combinatorial informatics
approach, the methodological dissection of
highly complex protein networks to reveal functional keystone factors is effi
ciently achieved. This form of practical informatics development represents
the next wave of biological interpretation of datasets of ever increasing volume.
In addition to tackling individual datasets of ever increasing volume, the need
for advanced data separation and classification of multiple simultaneous datasets
has also driven the creation of advanced platforms such as higher-order
Venn diagram analyzers.
16
The recently developed VENNTURE application
(
http://www.nia.nih.gov
) facilitates the ability of experimenters to manage
and investigate up to six parallel large datasets (
Fig. 17.1
A). To distinguish spe
cific
b
-arrestin signaling, it may be vital to correlate specific gene/protein
expression signatures in multiple diverse cell or tissue environments. The abil
ity to perform this over at least six different data streams may greatly accelerate
the discovery of a generic
b
-arrestin-specific signaling paradigm. Applications
such as VENNTURE may be able to revolutionize functional signaling anal
ysis by developing basic scientific approaches, that is, Venn diagram analysis.
Analytical advances are also currently under way at the other end of the com
plexity spectrum of data analysis, for example, the recently developed field
of
data texturization.
Data texturization applications such as the NIH
attempt to explore previously cryptic nuances in highly complex datasets by
creating three-dimensional proxies of the data possessing various structural
or chromatic properties. Data texturization may represent one of the ultimate
forms of signaling data analysis that may allow the effective combination
of human user investigation and machine-based learning algorithms to
uncover idiosyncratic and discriminatory patterns within hypercomplex cell
signaling data.
Search WWH ::
Custom Search