Supervised reconstruction of biological networks with local models

By Kevin Bleakley, Gérard Biau, Jean-Philippe Vert. ISMB/ECCB, 2007

575 citations recorded by the source corpus.

MOTIVATION Inference and reconstruction of biological networks from heterogeneous data is currently an active research subject with several important applications in systems biology. The problem has been attacked from many different points of view with varying degrees of success. In particular, predicting new edges with a reasonable false discovery rate is highly demanded for practical applications, but remains extremely challenging due to the sparsity of the networks of interest. RESULTS While most previous approaches based on the partial knowledge of the network to be inferred build global models to predict new edges over the network, we introduce here a novel method which predicts whether there is an edge from a newly added vertex to each of the vertices of a known network using local models. This involves learning individually a certain subnetwork associated with each vertex of the known network, then using the discovered classification rule associated with only that vertex to predict the edge to the new vertex. Excellent experimental results are shown in the case of metabolic and protein-protein interaction network reconstruction from a variety of genomic data. AVAILABILITY An implementation of the proposed algorithm is available upon request from the authors.

Subfields: Artificial Intelligence, Molecular Biology

View Paper: https://www.semanticscholar.org/paper/ce09e12e6c05fd7f63d23af63ef581577e01d803

How researchers see this paper

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

In Artificial Intelligence

In Molecular Biology

Citations in the corpus

Semi-supervised Penalized Output Kernel Regression for Link Prediction (2011, unknown): “The ﬁrst three lines come from (Bleakley et al., 2007): em stands for em projection method (Tsuda et al., 2003), Pkernel for tensor product pairwise kernel with SVM (Ben-Hur & Noble, 2005) and local for local models with SVM (Bleakley et al., 2007).”
l p -Norm Multiple Kernel Learning (2011, methodology): “Chapelle et al. (2002) and Bousquet and Herrmann (2002) minimize estimates of the generalization error of support vector machines (SVMs) using a gradient descent algorithm over the set of parameters.”
Semi-supervised multi-task learning for predicting interactions between HIV-1 and human proteins (2010, methodology): “(Bleakley et al., 2007) that uses a local model to allow for flexible modeling of subgroups of interactions.”
A Transfer Learning Approach and Selective Integration of Multiple Types of Assays for Biological Network Inference (2010, methodology): “We use only high confidence edges as in previous studies (Kato et al., 2005; Bleakley et al., 2007), because they are supported by multiple experiments.”
Input Output Kernel Regression (2016, background): “• Local (Bleakley et al., 2007): a local model is built for each protein in order to learn the subnetwork associated to each protein and these models are then combined together.”
Machine learning approaches and databases for prediction of drug–target interaction: a survey paper (2020, methodology): “target using local models trained on known drugs and targets [98, 101, 107, 108].”
Supervised prediction of drug–target interactions using bipartite local models (2009, methodology): “Versions of local models have previously been very successful in learning and predicting protein-protein interaction and metabolic networks (Bleakley et al. , 2007) and regulatory networks (Mordelet and Vert, 2008).”
Similarity-based machine learning methods for predicting drug-target interactions: a brief review (2014, background): “BLM extends the idea, called local models , which was proposed in [27, 57, 58].”
Accounting, decisions and promises (2016, background): “In their account, Akrich, Callon, and Latour (2002a; 2002b) argue that innovation is never adopted and always adapted.”