Kernel methods for predicting protein-protein interactions

By Asa Ben-Hur, William Stafford Noble. Intelligent Systems in Molecular Biology, 2005

638 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/09f05bb4acfb1b58fa1173c554cf7bf205672c24

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Molecular Biology

Citations in the corpus

Studying protein^protein interaction networks: a systems view on diseases (2012, methodology): “Besides the methods based on a single data source, many bioinformaticians use multiple data sources, including Bayesian network approach [15], kernels method [16], probabilistic decision tree approach [17] and the hybrid approach [18].”
A kernel method for unsupervised structured network inference (2009, methodology): “Previous kernel approaches to network inference exhibit at least one of the following three properties ( Ben-Hur & Noble, 2005; Vert et al., 2007):”
Minimum curvilinearity to enhance topological prediction of protein interactions by network embedding (2013, methodology): “For completeness, we compared ncMCE-SP with FSW and CDD, two of the most efficient node neighbourhood techniques (You et al., 2010). ncMCE-SP ranked first, with a notable improvement, in the first two networks (Ben-Hur and Noble 2005; Chen et al. 2006; Supplementary Fig.”
Support vector machines (2008, positive): “, 2002), protein sub-cellular localization Ong and Zien (2008), to detect protein protein interactions (Ben-Hur and Noble, 2005) to analyze phylogenetic trees Vert (2002).”
Protein-Protein Interaction Detection: Methods and Analysis (2014, background): “Kernelmethods are hybridmethodswhich use a combination of properties like protein sequences, gene ontologies, and so forth [41].”
A Transfer Learning Approach and Selective Integration of Multiple Types of Assays for Biological Network Inference (2010, methodology): “TPPK is based on a tensorization of the original feature space, whereas MLPK is derived from convex optimization of a distance metric learning problem.”
Computational approaches for detecting protein complexes from protein interaction networks: a survey (2010, methodology): “integrate weighted genomic features in [50] and mapping different features into high dimensional vector space in [106]).”
Semi-supervised Penalized Output Kernel Regression for Link Prediction (2011, neutral): “…posterior probabilities of links while kernel or similarity based methods take beneﬁt from the versatility of kernels to encode various structured knowledge in the input space as well as in the output space ( Ya-Appearing manishi et al., 2004; Ben-Hur & Noble, 2005; Geurts et al., 2006; 2007a;b).”
Interactome: gateway into systems biology. (2005, methodology): “Computational prediction efforts have utilized conserved protein and DNA sequences, functional annotations based on GO, co-localization or homologous interactions in other species, and are complemented by mathematical modeling employing strategies such as Bayesian networks (75–78,80,119–122).”
Learning Kernels with Random Features (2016, background): “In the supervised setting, one assumes a base class or classes of kernels and either uses heuristic rules to combine kernels [2] [23], optimizes structured (e.g. linear, nonnegative, convex) compositions of the kernels with respect to an alignment metric [9] [16] [20] [28], or jointly optimizes…”
Semi-supervised multi-task learning for predicting interactions between HIV-1 and human proteins (2010, methodology): “(2004), kernel-based methods from Ben-Hur and Noble (2005) and Yamanishi et al.”
Input Output Kernel Regression (2016, methodology): “The pairwise kernel method (Ben-Hur and Noble, 2005) was not considered here because this method requires to define a Gram matrix between pairs of nodes, which raises some practical issues in terms of computation time and storage.”