Biology Reference
In-Depth Information
FIGURE 8.4 The cellular response to chemical stress. Biological processes and protein complexes frequently associated with chemical stress.
(A) Each node represents a gene ontology (GO) biological process or protein complex that is significantly enriched in at least 1% of ~3500 HIPHOP
profiles (FDR 0.1 by gene set enrichment analysis, [68] ). The size of a node corresponds to the number of profiles in which the gene category is
enriched, with larger nodes indicating more frequent enrichment. The width of an edge is proportional to the level of gene overlap between the two
connected categories (i.e., gene sets). The color of a node shows its cluster membership, where clustering is based on the level of overlap between
categories and thus groups together related categories. Each cluster is labeled with a general name that describes the member nodes. For the 'ergosterol
biosynthesis' cluster, we show the top four most significantly enriched ring assemblies amongst the small molecules that are associated with the cluster,
via enrichments in their profiles. (B) GO biological processes and protein complexes that are frequently associated with chemical stress, either alone or in
combination. All considered processes and complexes were clustered based on the level of overlap between categories, and each resulting cluster is
represented as a class of processes and complexes (pink nodes). A blue-bordered node is associated with a set of at least five small molecules with
HIPHOP profiles that are significantly enriched (FDR 0.1 by gene set enrichment analysis, [68] ) for processes/complexes only in the classes to which
the node is connected (by pink edges). The size of a blue-bordered node is proportional to the number of small molecules associated with it, with larger
nodes indicating more molecules. By extension, the sizes of these nodes indicate the prevalence of particular enrichment signatures. Two blue-bordered
nodes are connected by a blue edge if at least 10% of the associated between-set small molecule pairs exhibit significant co-inhibition (p
0.01).
Interestingly, blue edges only occur between small molecule sets enriched for amino acid and nucleotide metabolism processes. Finally, each blue-
bordered node contains a nested co-inhibition network of the associated small molecule set, where nodes represent molecules and edges represent
significant co-inhibition (p
0.01). The main network can be mined to identify chemical classes that target or at least affect multiple processes, and to
potentially identify functional relationships between different processes. These functional relationships may only become apparent under specific
perturbation conditions. The network was made using Cytoscape software [69] .
Search WWH ::




Custom Search