Open Access Open Badges Method

A computational framework for boosting confidence in high-throughput protein-protein interaction datasets

Raghavendra Hosur1, Jian Peng12, Arunachalam Vinayagam3, Ulrich Stelzl4, Jinbo Xu2, Norbert Perrimon35, Jadwiga Bienkowska6* and Bonnie Berger17*

Author Affiliations

1 Computer Science and Artificial Intelligence Laboratory, 32 Vassar Street, MIT, Cambridge, MA 02139, USA

2 Toyota Technological Institute, 6045 S. Kenwood Ave, Chicago, IL 60637, USA

3 Department of Genetics, 77 Avenue Louis Pasteur, Harvard Medical School, Boston, MA 02115, USA

4 Otto-Warburg Laboratory, Ihnestraβe 63-73, Max Planck Institute for Molecular Genetics, Berlin D14195, Germany

5 Howard Hughes Medical Institute, 20 Shattuck Street, Boston, MA 02115, USA

6 Computational Biology group, Biogen Idec, 14 Cambridge Center, Cambridge, MA 02142, USA

7 Department of Mathematics, 77 Massachusetts Avenue, MIT, Cambridge, MA 02139, USA

For all author emails, please log on.

Genome Biology 2012, 13:R76  doi:10.1186/gb-2012-13-8-r76

Published: 31 August 2012


Improving the quality and coverage of the protein interactome is of tantamount importance for biomedical research, particularly given the various sources of uncertainty in high-throughput techniques. We introduce a structure-based framework, Coev2Net, for computing a single confidence score that addresses both false-positive and false-negative rates. Coev2Net is easily applied to thousands of binary protein interactions and has superior predictive performance over existing methods. We experimentally validate selected high-confidence predictions in the human MAPK network and show that predicted interfaces are enriched for cancer -related or damaging SNPs. Coev2Net can be downloaded at webcite.