Fwd: CSHL Meeting Abstract Submission

Thank you for submitting your abstract for Systems Biology: Global
Regulation of Gene Expression 2014 .

Comparative network analysis of ENCODE and modENCODE data

Mark Gerstein1,2, Koon-Kiu Yan1,2, modENCODE/ENCODE transcriptome group1

1Yale University, Program of Computational Biology and Bioinformatics,
New Haven, CT, 2Yale University, Molecular Biophysics and
Biochemistry, New Haven, CT

The ENCODE and modENCODE consortia have generated a resource
containing large amounts of transcriptomic data, extensive mapping of
chromatin states, as well as the binding locations of over 300
transcription-regulatory factors for human, worm and fly. We performed
extensive data integration by constructing genome-wide co-expression
networks and transcriptional regulatory networks, revealing
fundamental principles of transcription and network organization that
are conserved across the three highly divergent animals. First, we
developed a novel cross-species clustering algorithm to integrate the
co-expression networks of the three species, resulted at conserved
modules shared between the organisms. These modules are enriched in
developmental genes and exhibited hourglass behavior. They were then
used to align the stages in worm and fly development, finding the
normal embryo-to-embryo and larvae-to-larvae pairings in addition to a
novel pairing between worm embryo and fly pupae. Second, we defined a
new score to quantify the degree of hierarchy (preponderance of
downward information flow) of a network, and developed a global
optimization algorithm to compare the hierarchical organization of the
three species. We found that, despite extensive rewiring of binding
targets, high-level organization principles like the hierarchical
structures are conserved across three species. Finally, we found the
gene expression levels in the organisms, both coding and non-coding,
can be predicted consistently based on their upstream histone marks.
In fact, a "universal model" with a single set of cross-organism
parameters can predict expression level for both protein coding genes
and ncRNAs. The algorithms introduced by this study can be easily
applied to other model datasets such as those from yeast.

i0cshsb