have developed general framework for

Posted by Ellie on Mar 17, 2009 in ISMB Conference | 0 comments

Parts ofthis work have been published in BIBE PKDD2006, LinkKDD and ISMB have also examined the use ofensemble clustering for this purpose, with successful results. Post Graduate Research Associate present Supervisor Dr. Srinivasan Parthasarathy Relevant Projects Functional Clustering ofInteraction Networks The objective here is to extract usefulmodules or clusters from realworld interaction networks.

InProteinProtein interaction PPI networks, the discovery of keyfunctional modules can help understand the functions of proteins andalso aid in predicting the function of unknown unannotated proteins. Traditional clusteringgraph partitioning algorithms have not performedwell in this task due to the presence of noisy false positiveinteractions scalefree topology, and multifaceted hub nodes. I have developed efficient techniques focusing on the topologicalproperties of these networks to eradicate noise and discoverfunctionally relevant clusters. Post Graduate Course Instructor Introduction to Computer Science CSE100.

Asur, In theProceedings of the 13th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining SIGKDD, Raman, Parthasarathy and Asur, Asur, InProteinProtein interaction PPI networks, the discovery of keyfunctional modules can help understand the functions of proteins andalso aid in predicting the function of unknown unannotated proteins. Traditional clusteringgraph partitioning algorithms have not performedwell in this task due to the presence of noisy false positiveinteractions scalefree topology, and multifaceted hub nodes. I have developed efficient techniques focusing on the topologicalproperties of these networks to eradicate noise and discoverfunctionally relevant clusters.

Parthasarathy, have also examined the evolutionary behavior of these neighborhoods over time. Post Graduate Research Associate present Supervisor Dr. Srinivasan Parthasarathy Relevant Projects Functional Clustering ofInteraction Networks The objective here is to extract usefulmodules or clusters from realworld interaction networks. An Ensemble Approach for ClusteringScaleFree Graphs. Wang, Effective Preprocessing Strategies forFunctional Clustering of ProteinProtein Interactions Network.

Post Graduate Research Associate present Supervisor Dr. Srinivasan Parthasarathy Relevant Projects Functional Clustering ofInteraction Networks The objective here is to studyevolving realworld interaction networks, such as social networks, WWWnetworks and biological networks geneexpression timeseries networks.Identifying the portions of the network that are changing,characterizing the type of change, predicting future events linkprediction, and developing generic models for evolving networks arecritical challenges that have looked to address.
Tags: algorithms, Computer Science, evolution