Biology Reference
In-Depth Information
(“stemness”), and those for specific cell type differentiation. A practical
value in knowing the identity of these genes and the pathways they
regulate is the use of such knowledge to advance our ability to isolate,
maintain, and differentiate ES cells.
Network Predictions: Transcriptome Analysis of ES
and Embryonic Cells
Earlier attempts using methods such as differential screens and sub-
tractive hybridization uncovered only a handful of genes preferentially
expressed in ES cells and, by implication, important for their unique
properties. With the advent of the idea of spotting multiple probes on
microarray chips [11], it became possible to interrogate RNAs simulta-
neously with a large collection of gene probes in a multiplex fashion for
the transcriptome expressed in ES cells and during differentiation.
Such an approach immediately revealed that the number of genes that
may be important for maintaining ES cell states could be in the hun-
dreds [12-14]. Data sets combined from studies performed by several
groups using similar arrays further provided the opportunity to inter-
sect databases to find common elements [14]. This comparative
analysis revealed a set of genes that are found by all studies to be pref-
erentially expressed in ES cells compared to other cell types. Because
the list includes genes that are known to be important for stemness
properties of ES cells, such as Oct-4/Pou5f1 and Nanog (figure 10.3), by
implication some or many of the genes in this list would likely be
involved in ES cell growth and development.
However, an integrated picture of how these genes work together to
determine ES stemness has not emerged and remains to be established. To
demonstrate how such lists of genes may be further utilized to reveal
biological meaning, we analyzed the 332 “ES genes” shown in figure 10.3
using Ingenuity's Pathway analysis software ( http://www.ingenuity.
com) and Pathway Assist ( http://www.ariadnegenomics.com). These
programs identify and extend connections between a given set of genes
using the enormous wealth of information buried in scientific literature.
These two approaches differ in the way their respective knowledge
bases are developed. Ingenuity's knowledge base is manually indexed,
increasing its accuracy, but this means that it covers less of the scientific
literature (figure 10.4a). Meanwhile, Pathway Assist (figure 10.4b) uses
an automated natural language processor to gain knowledge from the
literature rendering a noisy knowledge base with reduced accuracy,
but gains a more exhaustive coverage [15].
Interestingly, despite differences in the method of knowledge base
development between the two programs, very similar major network
pathways were revealed when the 332 ES genes were analyzed (figures
10.4a and 10.4b). Each of these networks has a hub-like node and here
we show the integration of the p53 tumor suppressor, c-myc and c-fos
Search WWH ::




Custom Search