Updated project metadata. ~500,000 protein interactions have been identified in human but most binding affinities remain unknown, limiting quantitative investigation of interactome-function relationships. This gap can be filled by systematically measuring the affinities of minimal interacting protein fragments, the building blocks of interactomes. As a proof of principle, we experimentally measured 65,000 affinities involving 266 human PDZ domains (practically the complete "PDZome"), 424 human PDZ-Binding Motifs (~10% of the "PBMome") and 24 viral motifs. We determined binding constants for ~18,000 interactions that passed the detection threshold. Independent AP-MS experiments and database surveys demonstrated the strong agreement of quantitative fragmentomics with full-length protein interactomics. We evidence hot spots for viral interference, which allow a viral PBM, such as that of HPV-E6 oncoprotein, to impact the cell proteome way beyond its immediate binders. We propose to distinguish interactomes and complexomes, viewed as intrinsic and extrinsic properties linked by the law of mass action.