Detection of protein-protein interactions is usually a balance between specificity (background reduction) and affinity (detection of weak interactions). Recently, the introduction of stable isotope labeling in mass spectromet-ric approaches allows one to distinguish specific from unspecific interaction partners in an immunoprecipitate or pull-down experiment. This principle has enabled detection of weak binders in the presence of background proteins, and it has led to a decrease in false positive detections (Blagoev et al. 2003; Gruhler et al. 2005).
However, there are still some problems associated with large-scale experimental interaction mapping: in most large-scale experimental datasets, false negative rates range up to 90% and false positive rates are still predicted to be higher than 50%, even in highly filtered datasets (von Mering et al. 2002). Bioinformatic methods will increasingly be necessary to contribute to the validation and prediction of protein interactions.
Was this article helpful?