Display options
Share it on

BMC Syst Biol. 2018 Apr 11;12:9. doi: 10.1186/s12918-018-0535-4.

Annotating activation/inhibition relationships to protein-protein interactions using gene ontology relations.

BMC systems biology

Soorin Yim, Hasun Yu, Dongjin Jang, Doheon Lee

Affiliations

  1. Department of Bio and Brain Engineering, KAIST, 291 Daehak-ro, Yuseong-gu, Daejeon, 34141, Republic of Korea.
  2. Bio-Synergy Research Center, 291 Daehak-ro, Yuseong-gu, Daejeon, 34141, Republic of Korea.
  3. Department of Bio and Brain Engineering, KAIST, 291 Daehak-ro, Yuseong-gu, Daejeon, 34141, Republic of Korea. [email protected].
  4. Bio-Synergy Research Center, 291 Daehak-ro, Yuseong-gu, Daejeon, 34141, Republic of Korea. [email protected].

PMID: 29671402 PMCID: PMC5907154 DOI: 10.1186/s12918-018-0535-4

Abstract

BACKGROUND: Signaling pathways can be reconstructed by identifying 'effect types' (i.e. activation/inhibition) of protein-protein interactions (PPIs). Effect types are composed of 'directions' (i.e. upstream/downstream) and 'signs' (i.e. positive/negative), thereby requiring directions as well as signs of PPIs to predict signaling events from PPI networks. Here, we propose a computational method for systemically annotating effect types to PPIs using relations between functional information of proteins.

RESULTS: We used regulates, positively regulates, and negatively regulates relations in Gene Ontology (GO) to predict directions and signs of PPIs. These relations indicate both directions and signs between GO terms so that we can project directions and signs between relevant GO terms to PPIs. Independent test results showed that our method is effective for predicting both directions and signs of PPIs. Moreover, our method outperformed a previous GO-based method that did not consider the relations between GO terms. We annotated effect types to human PPIs and validated several highly confident effect types against literature. The annotated human PPIs are available in Additional file 2 to aid signaling pathway reconstruction and network biology research.

CONCLUSIONS: We annotated effect types to PPIs by using regulates, positively regulates, and negatively regulates relations in GO. We demonstrated that those relations are effective for predicting not only signs, but also directions of PPIs. The usefulness of those relations suggests their potential applications to other types of interactions such as protein-DNA interactions.

Keywords: Activation; Gene ontology; Inhibition; Protein-protein interaction

References

  1. Comput Biol Med. 2014 Aug;51:198-204 - PubMed
  2. Nucleic Acids Res. 2017 Jan 4;45(D1):D362-D368 - PubMed
  3. BMC Med Genomics. 2015;8 Suppl 2:S2 - PubMed
  4. Biochem Pharmacol. 2011 Nov 15;82(10):1269-76 - PubMed
  5. Cell Metab. 2014 Jun 3;19(6):967-80 - PubMed
  6. Proc Natl Acad Sci U S A. 2001 Mar 27;98 (7):3802-7 - PubMed
  7. Nat Med. 2013 Feb;19(2):179-92 - PubMed
  8. Nat Rev Clin Oncol. 2015 Aug;12(8):445-64 - PubMed
  9. Mol Cell. 2012 Mar 9;45(5):619-28 - PubMed
  10. Sci Signal. 2011 Sep 6;4(189):rs8 - PubMed
  11. Nat Methods. 2014 Jan;11(1):94-9 - PubMed
  12. Bioinformatics. 2003 Jul 1;19(10):1275-83 - PubMed
  13. J Biol Chem. 2011 Feb 25;286(8):5956-66 - PubMed
  14. Nat Rev Cancer. 2014 Nov;14(11):736-46 - PubMed
  15. Nat Rev Cancer. 2013 Jan;13(1):11-26 - PubMed
  16. J Biomed Inform. 2002 Oct-Dec;35(5-6):352-9 - PubMed
  17. Mol Syst Biol. 2015 Dec 17;11(12 ):848 - PubMed
  18. Nucleic Acids Res. 2017 Jan 4;45(D1):D369-D379 - PubMed
  19. PLoS Comput Biol. 2014 Sep 25;10(9):e1003814 - PubMed
  20. J Natl Cancer Inst. 2014 Jan;106(1):djt356 - PubMed
  21. Neuron. 2013 May 22;78(4):644-57 - PubMed
  22. Nucleic Acids Res. 2015 Jan;43(Database issue):D1049-56 - PubMed
  23. Nucleic Acids Res. 2017 Jan 4;45(D1):D353-D361 - PubMed
  24. Sci Rep. 2016 Nov 07;6:36453 - PubMed
  25. BMC Syst Biol. 2016 Jan 11;10 Suppl 1:2 - PubMed
  26. PLoS One. 2013 Nov 18;8(11):e79606 - PubMed

MeSH terms

Publication Types