|
2.2 Comparative and
integrative analysis of 'omics' data.
Responsible:
Benedikt Brors, DKFZ, Heidelberg.
Background:
One of the strengths
of the NGFN2 approach is the diversity of data-types that are being
generated in numerous disease areas. These include the abundance of
molecules or transcriptional units, as well as mutations either on
genomic/transcriptomic or transcriptomic/proteomic levels. Despite the
fact that additional high-value knowledge could be gained by
integrating all available data in an analysis, efficient methods for
such analyses are still not available. This sub-project intends to
fill this gap by systematically comparing data on all three levels. We
will study their particular strengths and weaknesses, and develop
methods for the simultaneous assessment especially in terms of their
associations with disease phenotypes.
Planned Work:
We will systematically
explore concordance as well as discrepancies between data
investigating genomic, proteomic and transcriptomic changes in samples
of diseased tissue, and develop methods for integrating them
simultaneously into analyses. In particular, we will:
-
Check for correlations among
proteomic and transcriptomic data.
-
Develop means for integrating
heterogeneous data into a single analysis.
-
Develop consistency checks for the
data during integration.
-
Develop methods to predict genomic
changes from transcriptomic analyses.
-
Conduct a prototypic analysis on a
complete vertical 'omics' data set.
Publications
Warnat P, Eils R, Brors B.
Cross-platform analysis of cancer microarray data improves gene
expression based classification of phenotypes. BMC Bioinformatics. 2005 Nov 4;6(1):265.
Shahi P, Loukianiouk S, Bohne-Lang A, Kenzelmann M, Küffer S,
Maertens S, Eils R, Gröne HJ, Gretz N, Brors B. Argonaute-a database for gene regulation by
mammalian microRNAs. Nucleic Acids Res. 2006 Jan 1;34. |
|
