MIF provides only a high-level description of experiments. Other file formats (including these developed by HUPO-PSI) provide more detailed description. It seems reasonable to define some formal mechanism of referencing these within MIF - possibly by adding one or more CV terms to be used as xref type
lukasz