Inductive matrix completion for predicting gene–disease associations

Motivation: Most existing methods for predicting causal disease genes rely on specific type of evidence, and are therefore limited in terms of applicability. More often than not, the type of evidence available for diseases varies—for example, we may know linked genes, keywords associated with the disease obtained by mining text, or co-occurrence of disease symptoms in patients. Similarly, the type of evidence available for genes varies—for example, specific microarray probes convey information only for certain sets of genes. In this article, we apply a novel matrix-completion method called Inductive Matrix Completion to the problem of predicting gene-disease associations; it combines multiple types of evidence (features) for diseases and genes to learn latent factors that explain the observed gene–disease associations. We construct features from different biological sources such as microarray expression data and disease-related textual data. A crucial advantage of the method is that it is inductive; it can be applied to diseases not seen at training time, unlike traditional matrix-completion approaches and network-based inference methods that are transductive. Results: Comparison with state-of-the-art methods on diseases from the Online Mendelian Inheritance in Man (OMIM) database shows that the proposed approach is substantially better—it has close to one-in-four chance of recovering a true association in the top 100 predictions, compared to the recently proposed Catapult method (second best) that has <15% chance. We demonstrate that the inductive method is particularly effective for a query disease with no previously known gene associations, and for predicting novel genes, i.e. genes that are previously not linked to diseases. Thus the method is capable of predicting novel genes even for well-characterized diseases. We also validate the novelty of predictions by evaluating the method on recently reported OMIM associations and on associations recently reported in the literature. Availability: Source code and datasets can be downloaded from http://bigdata.ices.utexas.edu/project/gene-disease. Contact: naga86@cs.utexas.edu

[1]  C. Carter.  Mendelian Inheritance in Man , 1967 .

[2]  J. Rashbass.  Online Mendelian Inheritance in Man. , 1995, Trends in genetics : TIG.

[3]  D. Valle,et al.  Online Mendelian Inheritance In Man (OMIM) , 2000, Human mutation.

[4]  A. F. Scott,et al.  OMIM: Online Mendelian Inheritance in Man , 2002 .

[5]  Kara Dolinski,et al.  Saccharomyces Genome Database (SGD) provides secondary gene annotation using the Gene Ontology (GO) , 2002, Nucleic Acids Res..

[6]  S. Amladi,et al.  Online Mendelian Inheritance in Man 'OMIM'. , 2003, Indian journal of dermatology, venereology and leprology.

[7]  David J. Porteous,et al.  Speeding disease gene discovery by sequence based candidate prioritization , 2005, BMC Bioinformatics.

[8]  N. Campbell.  Genetic association database , 2004, Nature Reviews Genetics.

[9]  G. Bell,et al.  GEISHA, a whole‐mount in situ hybridization gene expression screen in chicken embryos , 2004, Developmental dynamics : an official publication of the American Association of Anatomists.

[10]  C. Ouzounis,et al.  Genome-wide identification of genes likely to be involved in human genetic disease. , 2004, Nucleic acids research.

[11]  Shinichi Morishita,et al.  SCMD: Saccharomyces cerevisiae Morphological Database , 2004, Nucleic Acids Res..

[12]  Kimberly Van Auken,et al.  WormBase: a comprehensive data resource for Caenorhabditis biology and genomics , 2004, Nucleic Acids Res..

[13]  Yongjin Li,et al.  Discovering disease-genes by topological features in human protein-protein interaction network , 2006, Bioinform..

[14]  G. Vriend,et al.  A text-mining analysis of the human phenome , 2006, European Journal of Human Genetics.

[15]  Monte Westerfield,et al.  The Zebrafish Information Network: the zebrafish model organism database , 2005, Nucleic Acids Res..

[16]  James Bennett,et al.  The Netflix Prize , 2007 .

[17]  Judith A. Blake,et al.  The mouse genome database (MGD): new features facilitating a model system , 2006, Nucleic Acids Res..

[18]  Kriston L. McGary,et al.  Broad network-based predictability of Saccharomyces cerevisiae gene loss-of-function phenotypes , 2007, Genome Biology.

[19]  P. Robinson,et al.  Walking the interactome for prioritization of candidate disease genes. , 2008, American journal of human genetics.

[20]  Michael Q. Zhang,et al.  Network-based global inference of human disease genes , 2008, Molecular systems biology.

[21]  Robert P. St.Onge,et al.  The Chemical Genomic Portrait of Yeast: Uncovering a Phenotype for All Genes , 2008, Science.

[22]  P. Provero,et al.  Functional Annotation and Identification of Candidate Disease Genes by Computational Analysis of Normal Tissue Gene Expression Data , 2008, PloS one.

[23]  Tanya Z. Berardini,et al.  The Arabidopsis Information Resource (TAIR): gene structure and function annotation , 2007, Nucleic Acids Res..

[24]  David Osumi-Sutherland,et al.  FlyBase: enhancing Drosophila Gene Ontology annotations , 2008, Nucleic Acids Res..

[25]  E. Snitkin,et al.  Genome-wide prioritization of disease genes and identification of disease-disease associations from an integrated human functional linkage network , 2009, Genome Biology.

[26]  Roded Sharan,et al.  A Network-Based Method for Predicting Disease-Causing Genes , 2009, J. Comput. Biol..

[27]  Jing Chen,et al.  ToppGene Suite for gene list enrichment analysis and candidate gene prioritization , 2009, Nucleic Acids Res..

[28]  Jagdish Chandra Patra,et al.  Genome-wide inferring gene-phenotype relationship by walking on the heterogeneous network , 2010, Bioinform..

[29]  A. Barabasi,et al.  Network medicine : a network-based approach to human disease , 2010 .

[30]  Roded Sharan,et al.  Associating Genes and Protein Complexes with Disease via Network Propagation , 2010, PLoS Comput. Biol..

[31]  Jean-Philippe Vert,et al.  ProDiGe: Prioritization Of Disease Genes with multitask machine learning from positive and unlabeled examples , 2011, BMC Bioinformatics.

[32]  N. Krogan,et al.  Phenotypic Landscape of a Bacterial Cell , 2011, Cell.

[33]  F. Piano,et al.  A High-Resolution C. elegans Essential Gene Network Based on Phenotypic Profiling of a Complex Tissue , 2011, Cell.

[34]  E. Marcotte,et al.  Prioritizing candidate disease genes by network-based boosting of genome-wide association data. , 2011, Genome research.

[35]  R. Piro,et al.  Computational approaches to disease‐gene prediction: rationale, classification and successes , 2012, The FEBS journal.

[36]  Y. Moreau,et al.  Computational tools for prioritizing candidate genes: boosting disease gene discovery , 2012, Nature Reviews Genetics.

[37]  Bart De Moor,et al.  An unbiased evaluation of gene prioritization tools , 2012, Bioinform..

[38]  Nagarajan Natarajan,et al.  Learning with Noisy Labels , 2013, NIPS.

[39]  Jaime G. Carbonell,et al.  Multitask learning for host–pathogen protein interactions , 2013, Bioinform..

[40]  Inderjit S. Dhillon,et al.  Provable Inductive Matrix Completion , 2013, ArXiv.

[41]  John O. Woods,et al.  Prediction and Validation of Gene-Disease Associations Using Methods Inspired by Social Network Analyses , 2013, PloS one.

[42]  Inderjit S. Dhillon,et al.  Large-scale Multi-label Learning with Missing Labels , 2013, ICML.