Premium
Identifying functional modules in the physical interactome of Saccharomyces cerevisiae
Author(s) -
Pu Shuye,
Vlasblom Jim,
Emili Andrew,
Greenblatt Jack,
Wodak Shoshana J.
Publication year - 2007
Publication title -
proteomics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.26
H-Index - 167
eISSN - 1615-9861
pISSN - 1615-9853
DOI - 10.1002/pmic.200600636
Subject(s) - interactome , saccharomyces cerevisiae , tandem affinity purification , computational biology , computer science , protein–protein interaction , yeast , interaction network , process (computing) , data mining , binary number , saccharomyces , biology , gene , mathematics , genetics , biochemistry , enzyme , arithmetic , affinity chromatography , operating system
Reliable information on the physical and functional interactions between the gene products is an important prerequisite for deriving meaningful system‐level descriptions of cellular processes. The available information about protein interactions in Saccharomyces cerevisiae has been vastly increased recently by two comprehensive tandem affinity purification/mass spectrometry (TAP/MS) studies. However, using somewhat different approaches, these studies produced diverging descriptions of the yeast interactome, clearly illustrating the fact that converting the purification data into accurate sets of protein–protein interactions and complexes remains a major challenge. Here, we review the major analytical steps involved in this process, with special focus on the task of deriving complexes from the network of binary interactions. Applying the Markov Cluster procedure to an alternative yeast interaction network, recently derived by combining the data from the two latest TAP/MS studies, we produce a new description of yeast protein complexes. Several objective criteria suggest that this new description is more accurate and meaningful than those previously published. The same criteria are also used to gauge the influence that different methods for deriving binary interactions and complexes may have on the results. Lastly, it is shown that employing identical procedures to process the latest purification datasets significantly improves the convergence between the resulting interactome descriptions.