In my talk I'll try to cover three topics.

A. Pathway based analysis of the hematopoetic transcriptome.

B. Network based analysis of the transcriptional regulatory network in cancer.

We present an approach to utilizing the structure of transcriptional regulatory networks to classify tumor samples in terms of transcription factor-gene activities instead of gene expression profiles. To explore state-dependent regulatory circuitries, we constructed estimations of normal and malignant cell regulatory networks by combining evidence of co-expression from microarray gene expression datasets with connectivity networks, represented by transcription factor-gene regulatory relationships based on sequence predictions and literature based evidence. Our results show that active links between regulating-regulated gene pairs, ignoring expression levels of individual genes, have the capacity to separate samples or cell states. This approach has the advantage of allowing the identification of key transcription factor-gene pairs with differential activity, for example between two types of leukemia. We further studied collective gene expression and network properties by evaluating the topological proximity of differentially expressed genes. Gene sets that optimally classify samples, such as poor prognosis node-negative breast cancers, exhibit a surprising degree of connectivity, and are concentrated in neighborhoods in the gene regulation networks. Taken together these approaches provide the opportunity to predict the locations of dysregulated sub-networks. The study of features of transcriptional networks, although they are currently approximations, can contribute to understanding deregulated processes associated with cancer.

C. identification of condition dependent gene transcriptional regulatory network in yeast Transcriptional regulatory networks are commonly represented by connectivity maps generated from gene expression and transcription factor (TF) binding data. However, these connectivity maps are often composites of several cell states, and thus fail to illustrate the active transcriptional regulatory network that actually exists during a given cell condition. We present an approach that simultaneously decomposes the composite transcriptional regulatory network to condition-specific networks. It also allows us predict additional TF-target gene regulatory interactions that are active in these condition specific networks.

 

Characterizing disease states from topological properties of transcriptional regulatory networks, Tuck, D.P., Kluger, H.M., Kluger, Y. , BMC Bioinformatics 2006, 7:236

Unraveling condition specific gene transcriptional regulatory networks in Saccharomyces cerevisiae ,Kim, H., Hu, W., Kluger, Y., BMC Bioinformatics 2006, 7:165

Lineage specificity of gene expression patterns, Y Kluger , DP Tuck, JT Chang , Y Nakayama , R Poddar, N Kohya, Z Lian, A Ben Nasr, R Halaban, DS Krause, X Zhang, PE Newburger, and SM. Weissman (2004) PNAS 101:6508:6513

Inter- and intra-combinatorial regulation by transcription factors and microRNAs Yiming Zhou, John Ferguson, Joseph T Chang, Yuval Kluger BMC Genomics 2007, 8:396 (30 October 2007)