ANALYSIS NATURE METHODS | VOL.10  NO.4  | APRIL 2013  | 307 The characterization of all protein complexes of human cells under defined physiological conditions using affinity purification–mass spectrometry (AP-MS) is a highly desirable step in the quest to understand the phenotypic effects of genomic information. However, such a challenging goal has not yet been achieved, as it requires reproducibility of the experimental workflow and high data consistency across different studies and laboratories. We systematically investigated the reproducibility of a standardized AP-MS workflow by performing a rigorous interlaboratory comparative analysis of the interactomes of 32 human kinases. We show that it is possible to achieve high interlaboratory reproducibility of this standardized workflow despite differences in mass spectrometry configurations and subtle sample preparation–related variations and that combination of independent data sets improves the approach sensitivity, resulting in even more-detailed networks. Our analysis demonstrates the feasibility of obtaining a high-quality map of the human protein interactome with a multilaboratory project. Proteins engage in a plethora of stable and dynamic interactions to form macromolecular assemblies, essential components of any cellular process. The recent progress in sensitivity, throughput and specificity of AP-MS techniques 1–8 makes this technology ideal for the large-scale analysis of human protein complexes 9–20 . Roughly one-tenth of the estimated 650,000 protein-protein inter- actions (PPIs) in human cells have been experimentally observed so far 21,22 , which suggests that the complete characterization of the human interactome, similarly to early genome sequencing projects, will require the collaborative effort of a larger scien- tific community. The success of such collaborative AP-MS efforts toward completing a reference data set for the human interaction proteome strictly depends on the specificity and robustness in the respective sample preparation and data acquisition methods. Some of these issues have been addressed for collecting binary protein-interaction information by the popular yeast two-hybrid system 23–25 ; however, no sufficient evidence has been available so far that the protocols used to characterize complexes by AP-MS would be robust enough to warrant the large-scale multilabora- tory analysis necessary for the entire proteome. By analyzing the current protein interaction databases, we found that only 25% of annotated interactions are reproduced in an independent experi- ment using the same human protein as a bait (Supplementary Note 1). This rather low value could be explained by the hetero- geneity of the experimental systems and processing pipelines. We conducted a large-scale comparative interlaboratory study in which 32 human protein kinases were analyzed by AP-MS in parallel in two different laboratories: CeMM Research Center for Molecular Medicine of the Austrian Academy of Sciences (CeMM) and ETH Zurich (ETHZ). Samples and data were exchanged at all steps of the procedure to measure interlaboratory reproducibility in sample preparation, AP-MS data acquisition and computation. Although the question of mass spectrometry data reproducibil- ity has been addressed before 26 , to our knowledge this report is the first study in which the interlaboratory reproducibility of the entire AP-MS workflow, from sample preparation to bioinfor- matic filtering, is inspected in a systematic fashion. Our analysis showed that careful execution of a robust suite of experimental protocols using the same biological system leads to very high reproducibility within an individual laboratory (98%) and high overall interlaboratory reproducibility (81%). We found that the biochemical component affected the inter- laboratory reproducibility slightly more than mass spectrom- etry; however, these discrepancies were effectively minimized by applying stringent bioinformatic filtering. The study also showed that merging the AP-MS data sets from the two labo- ratories improved the resolution of the protein interaction net- work without sacrificing confidence. This work demonstrates that large-scale analyses of the human protein interactome by multilaboratory efforts are technically feasible and likely to produce results that are more accurate and complete than those of single-laboratory projects. RESULTS Cell-line generation, sample preparation and MS analysis We selected 32 kinases with average basal expression levels from kinases identified in a previous mass spectrometry survey as Interlaboratory reproducibility of large-scale human protein-complex analysis by standardized AP-MS Markku Varjosalo 1,4 , Roberto Sacco 2,4 , Alexey Stukalov 2,4 , Audrey van Drogen 1 , Melanie Planyavsky 2 , Simon Hauri 1 , Ruedi Aebersold 1,3 , Keiryn L Bennett 2 , Jacques Colinge 2 , Matthias Gstaiger 1 & Giulio Superti-Furga 2 1 Department of Biology, Institute of Molecular Systems Biology, ETH Zurich, Zurich, Switzerland. 2 CeMM Research Center for Molecular Medicine of the Austrian Academy of Sciences, Vienna, Austria. 3 Faculty of Science, University of Zurich, Zurich, Switzerland. 4 These authors contributed equally to this work. Correspondence should be addressed to G.S.-F. (gsuperti@cemm.oeaw.ac.at) or M.G. (gstaiger@imsb.biol.ethz.ch). RECEIVED 27 JULY 2012; ACCEPTED 29 JANUARY 2013; PUBLISHED ONLINE 3 MARCH 2013; DOI:10.1038/NMETH.2400 npg © 2013 Nature America, Inc. All rights reserved.