Year: 2005
Type: dissertation
Abstract: Obtaining a reliable interaction data set describing the human interactome is a \nmilestone yet to be reached. The past few years has seen tremendous progress in \nelucidating the yeast interactome. Experimental approaches for obtaining large-scale \nprotein interaction data coupled with powerful computational methods for combining \nthese data sets and for predicting functional relations between genes have been successful \nin tackling the yeast interactome. The concerted development of visualization techniques \nand the progress in the field of network biology has provided us with tools to evaluate, \nanalyze, and interpret the interactome. \nAlthough techniques are being scaled to tackle mammalian genomes, as witnessed \nby the first protein interaction networks for fly and worm we are far from a complete map \nof the human interactome. Human genes create additional challenges due to molecular \ncomplexity, tissue specificity, and alternate splicing. It therefore becomes important to \nbuild well-annotated benchmarks and accuracy measures to evaluate new data. \nHere, we describe three methods that provide a framework to build a \ncomprehensive human interactome. We have developed a novel algorithm for predicting \nprotein interaction partners based on comparing the position of proteins in their \nrespective phylogenetic trees. We establish two tests of the accuracy of human protein \ninteraction data sets and integrate the small-scale human interaction data sets using a Log \nlikelihood framework. The benchmarks and the consolidated interaction set will provide a \nbasis for determining the quality of future large-scale human protein interaction assays. \nLastly, based on patterns of conserved co-expression of human gene pairs and their \northologs from 5 different organisms (A. thaliana, M. musculus, D. melanogaster, C. \nelegans, and Yeast) we predict protein interactions, and test them against the benchmarks \nestablished by us. By combining the existing interaction data sets, we build a network of \n61,974 interactions between 9,642 human proteins and cluster the network to show \nexamples representative of the quality of the interactions in the network. \nThe methods, benchmarks and the Log likelihood framework, we hope, would \nenable us to build a comprehensive human interactome.
Author: Arun Ramani
Institution: The University of Texas at Austin
Language: English
Subfield: Molecular Biology
Domain: Life Sciences
Open Access status: green
OpenAlex ID: W1808065866
