Identification of candidate disease genes by integrating Gene Ontologies and protein-interaction networks: case study of primary immunodeficiencies

Nucleic Acids Res. 2009 Feb;37(2):622-8. doi: 10.1093/nar/gkn982. Epub 2008 Dec 10.


Disease gene identification is still a challenge despite modern high-throughput methods. Many diseases are very rare or lethal and thus cannot be investigated with traditional methods. Several in silico methods have been developed but they have some limitations. We introduce a new method that combines information about protein-interaction network properties and Gene Ontology terms. Genes with high-calculated network scores and statistically significant gene ontology terms based on known diseases are prioritized as candidate genes. The method was applied to identify novel primary immunodeficiency-related genes, 26 of which were found. The investigation uses the protein-interaction network for all essential immunome human genes available in the Immunome Knowledge Base and an analysis of their enriched gene ontology annotations. The identified disease gene candidates are mainly involved in cellular signaling including receptors, protein kinases and adaptor and binding proteins as well as enzymes. The method can be generalized for any disease group with sufficient information.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Computational Biology / methods
  • Genes
  • Genetic Predisposition to Disease*
  • Humans
  • Immunologic Deficiency Syndromes / genetics*
  • Protein Interaction Mapping*
  • Proteins / genetics
  • Terminology as Topic


  • Proteins