FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2201703519> ?p ?o ?g. }

• W2201703519 abstract "High-throughput methods for detecting protein-protein interactions (PPI) have recently gained popularity. These rapid advances in technology have given researchers an initial global picture of protein interactions on a genomic scale. The usefulness of this understanding is, however, typically compromised by noisy data and intrinsic complexity of the biological system. In this dissertation, we attempt to solve some problems in effectively analyzing the data. Firstly, since there are lots of false positives in experimentally detected interactions, we propose a novel topological measurement to select reliable interactions from the noisy data. Our method is based on the small-world network property of the protein interaction network and generalizes purely local measures adopted previously. Based on our observation that the true positive interactions in protein complexes and tightly coupled networks demonstrate dense interactions, we propose to measure the significance of two proteins' co-existence in a dense network as an index of interaction reliability. Our topological measure also integrates the prior confidence of each data set. The experiments demonstrate that our measure can be used to identify reliable interactions and to predict potential interactions with improved performance. Meanwhile, we discovered two additional properties: namely, the short alternative path property and the local clustering of network property of the protein interaction network, which are generalizations of previously known protein interaction network properties. Secondly, we address the problem of effectively incorporating domain knowledge into the protein clustering process. Based on our analysis of the relationship of network topology and biological relevance, we propose a novel semi-supervised clustering algorithm suitable for the noisy protein interaction network. We choose to estimate the pairwise similarity between each protein pair and use this similarity as input to clustering algorithms. Therefore, it is not bounded to any specific clustering methods. We select topological features in the network and define a model to map these features to pairwise similarities. The known protein annotations are used to train the model. Using this model, we can estimate the pairwise similarity between each pair of proteins. Finally, normal unsupervised clustering algorithms can be applied using the similarity matrix. Since our similarity measure has already incorporated prior protein annotations, our algorithm can detect clusters with improved performance. Also, the unsupervised clustering algorithms we adopt maintain the explorative nature and therefore are capable of detecting new protein functional groups. Thirdly, we investigate the problem of protein complex detection. Protein complexes can be roughly considered as densely connected subgraphs in the network. The difficulties in this problem are caused by the fact that protein complexes may overlap with each other, i.e. containing shared proteins, and the protein interaction network contains a lot of noise. To overcome these difficulties, we propose a novel subgraph quality measure, and based on the measure, we propose a novel seed-refine algorithm. Our subgraph quality measure achieves two goals: (1) it provides a statistically meaningful combination of inside links, outside links and the size of the subgraph and, (2) it provides a statistically meaningful combination of the quality contribution of each vertex in the subgraph. Our seed-refine algorithm consists of a two-layer seeding heuristic to find good seeds and a novel subgraph refinement method that controls the overlap between subgraphs. Our algorithm allows to output overlapping subgraphs but methodologically makes it possible only when there is strong evidence to do so. Experiments confirm the effectiveness of our method." @default.
• W2201703519 created "2016-06-24" @default.
• W2201703519 creator A5013588572 @default.
• W2201703519 creator A5052507774 @default.
• W2201703519 date "2007-01-01" @default.
• W2201703519 modified "2023-09-27" @default.
• W2201703519 title "Graph-based analysis of protein-protein interaction data sets" @default.
• W2201703519 hasPublicationYear "2007" @default.
• W2201703519 type Work @default.
• W2201703519 sameAs 2201703519 @default.
• W2201703519 citedByCount "0" @default.
• W2201703519 crossrefType "journal-article" @default.
• W2201703519 hasAuthorship W2201703519A5013588572 @default.
• W2201703519 hasAuthorship W2201703519A5052507774 @default.
• W2201703519 hasConcept C104317684 @default.
• W2201703519 hasConcept C105795698 @default.
• W2201703519 hasConcept C111472728 @default.
• W2201703519 hasConcept C111919701 @default.
• W2201703519 hasConcept C114614502 @default.
• W2201703519 hasConcept C119857082 @default.
• W2201703519 hasConcept C124101348 @default.
• W2201703519 hasConcept C138885662 @default.
• W2201703519 hasConcept C158154518 @default.
• W2201703519 hasConcept C177264268 @default.
• W2201703519 hasConcept C17744445 @default.
• W2201703519 hasConcept C185592680 @default.
• W2201703519 hasConcept C189950617 @default.
• W2201703519 hasConcept C199360897 @default.
• W2201703519 hasConcept C199539241 @default.
• W2201703519 hasConcept C199845137 @default.
• W2201703519 hasConcept C22047676 @default.
• W2201703519 hasConcept C2780009758 @default.
• W2201703519 hasConcept C28225019 @default.
• W2201703519 hasConcept C33923547 @default.
• W2201703519 hasConcept C38764148 @default.
• W2201703519 hasConcept C41008148 @default.
• W2201703519 hasConcept C55105296 @default.
• W2201703519 hasConcept C55493867 @default.
• W2201703519 hasConcept C64869954 @default.
• W2201703519 hasConcept C73555534 @default.
• W2201703519 hasConcept C80444323 @default.
• W2201703519 hasConceptScore W2201703519C104317684 @default.
• W2201703519 hasConceptScore W2201703519C105795698 @default.
• W2201703519 hasConceptScore W2201703519C111472728 @default.
• W2201703519 hasConceptScore W2201703519C111919701 @default.
• W2201703519 hasConceptScore W2201703519C114614502 @default.
• W2201703519 hasConceptScore W2201703519C119857082 @default.
• W2201703519 hasConceptScore W2201703519C124101348 @default.
• W2201703519 hasConceptScore W2201703519C138885662 @default.
• W2201703519 hasConceptScore W2201703519C158154518 @default.
• W2201703519 hasConceptScore W2201703519C177264268 @default.
• W2201703519 hasConceptScore W2201703519C17744445 @default.
• W2201703519 hasConceptScore W2201703519C185592680 @default.
• W2201703519 hasConceptScore W2201703519C189950617 @default.
• W2201703519 hasConceptScore W2201703519C199360897 @default.
• W2201703519 hasConceptScore W2201703519C199539241 @default.
• W2201703519 hasConceptScore W2201703519C199845137 @default.
• W2201703519 hasConceptScore W2201703519C22047676 @default.
• W2201703519 hasConceptScore W2201703519C2780009758 @default.
• W2201703519 hasConceptScore W2201703519C28225019 @default.
• W2201703519 hasConceptScore W2201703519C33923547 @default.
• W2201703519 hasConceptScore W2201703519C38764148 @default.
• W2201703519 hasConceptScore W2201703519C41008148 @default.
• W2201703519 hasConceptScore W2201703519C55105296 @default.
• W2201703519 hasConceptScore W2201703519C55493867 @default.
• W2201703519 hasConceptScore W2201703519C64869954 @default.
• W2201703519 hasConceptScore W2201703519C73555534 @default.
• W2201703519 hasConceptScore W2201703519C80444323 @default.
• W2201703519 hasLocation W22017035191 @default.
• W2201703519 hasOpenAccess W2201703519 @default.
• W2201703519 hasPrimaryLocation W22017035191 @default.
• W2201703519 hasRelatedWork W1820726125 @default.
• W2201703519 hasRelatedWork W195502668 @default.
• W2201703519 hasRelatedWork W1966879929 @default.
• W2201703519 hasRelatedWork W1973880717 @default.
• W2201703519 hasRelatedWork W1997987084 @default.
• W2201703519 hasRelatedWork W2014313978 @default.
• W2201703519 hasRelatedWork W2023111454 @default.
• W2201703519 hasRelatedWork W2025023381 @default.
• W2201703519 hasRelatedWork W2028561556 @default.
• W2201703519 hasRelatedWork W2076723453 @default.
• W2201703519 hasRelatedWork W2210891675 @default.
• W2201703519 hasRelatedWork W2246497063 @default.
• W2201703519 hasRelatedWork W2406861998 @default.
• W2201703519 hasRelatedWork W2743418339 @default.
• W2201703519 hasRelatedWork W2977578381 @default.
• W2201703519 hasRelatedWork W2981750204 @default.
• W2201703519 hasRelatedWork W3022597916 @default.
• W2201703519 hasRelatedWork W3080851671 @default.
• W2201703519 hasRelatedWork W3100526240 @default.
• W2201703519 hasRelatedWork W37408155 @default.
• W2201703519 isParatext "false" @default.
• W2201703519 isRetracted "false" @default.
• W2201703519 magId "2201703519" @default.
• W2201703519 workType "article" @default.