Advertisement
Research Article

What Evidence Is There for the Homology of Protein-Protein Interactions?

  • Anna C. F. Lewis,

    Affiliations: Department of Statistics, University of Oxford, Oxford, United Kingdom, Systems Biology Doctoral Training Centre, University of Oxford, Oxford, United Kingdom

    X
  • Nick S. Jones,

    Affiliations: Department of Mathematics, Imperial College, London, United Kingdom, Department of Physics, University of Oxford, Oxford, United Kingdom, CABDyN Complexity Centre, University of Oxford, Oxford, United Kingdom, Oxford Centre for Integrative Systems Biology, University of Oxford, Oxford, United Kingdom

    X
  • Mason A. Porter,

    Affiliations: CABDyN Complexity Centre, University of Oxford, Oxford, United Kingdom, Oxford Centre for Industrial and Applied Mathematics, Mathematical Institute, University of Oxford, Oxford, United Kingdom

    X
  • Charlotte M. Deane mail

    deane@stats.ox.ac.uk

    Affiliations: Department of Statistics, University of Oxford, Oxford, United Kingdom, Oxford Centre for Integrative Systems Biology, University of Oxford, Oxford, United Kingdom

    X
  • Published: September 20, 2012
  • DOI: 10.1371/journal.pcbi.1002645

Reader Comments (1)

Post a new comment on this article

Are the datasets comparable?

Posted by peteruetz on 26 Oct 2012 at 05:37 GMT

Even if you take into account incompleteness, I would argue that current datasets are hardly comparable as they have been produced not only in different species, but also with different methods that produce non-overlapping data with a large false negative rate. Take a look at this paper and you will see what I mean:

http://www.nature.com/nme...

Also, based on this paper I would also argue that the yeast interactome is still far from complete (but probably has lots of false positives). For instance, there hasn't even been a serious attempt to map all membrane protein interactions (except for Miller et al 2005 in PNAS, I believe).

It is obviously difficult, but I wonder how this analysis would look like when only PPIs are considered that are known to be of some relevance. Note that I DO believe that there are lots of bona fide PPIs that are NOT relevant to the cell :)

No competing interests declared.