Abstract
A considerable number of soluble proteins in cells that biochemists try to analyze are difficult to handle because they seem to behave like sponges that 'suck up' many other proteins. We argue here that this behavior is commonly an artifact introduced by the experimenting scientist and that we need to study proteins like animals in the wild: they will only reveal many of their secrets when carefully observed in their largely undisturbed, natural environment. Computational studies that attempt to realistically model cellular protein networks must also factor in the diverse protein habitats to be found in cells.